AS/400e series IBM

Performance Tools for AS/400

Version 4

SC41-5340-00
AS/400e series IBM
Performance Tools for AS/400
Version 4

SC41-5340-00
 Note

Before using this information and the product it supports, be sure to read the general information under “Notices” on page xiii.

First Edition (February 1998)

| This edition applies to the licensed programs IBM Performance Tools for AS/400 (Program 5769-PT1), Version 4 Release 2 Modifica-
| tion 0; IBM Operating System/400 (Program 5769-SS1), Version 4 Release 2 Modification 0, and to all subsequent releases and
| modifications until otherwise indicated in new editions. This edition applies only to reduced instruction set computer (RISC) systems.

 Copyright International Business Machines Corporation 1998. All rights reserved.

Note to U.S. Government Users — Documentation related to restricted rights — Use, duplication or disclosure is subject to
restrictions set forth in GSA ADP Schedule Contract with IBM Corp.
 Contents
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Programming Interface Information . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

About Performance Tools (SC41-5340) . . . . . . . . . . . . . . . . . . . . . . xvii

Who should read this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
AS/400 Operations Navigator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Prerequisite and related information . . . . . . . . . . . . . . . . . . . . . . . . xviii
| Information available on the World Wide Web . . . . . . . . . . . . . . . . . . . xix
| How to send your comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix

| Summary of Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

| Chapter 1. Introduction to Performance Tools . . . . . . . . . . . . . . . . . 1-1

| Why Manage Performance? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
| Performance Capabilities—OS/400 and Performance Tools . . . . . . . . . . . 1-2
| OS/400 Has Performance Management Basics . . . . . . . . . . . . . . . . . 1-2
| Performance Tools Builds on OS/400 Function . . . . . . . . . . . . . . . . . 1-2
| Manager Features and Agent Features . . . . . . . . . . . . . . . . . . . . . 1-3
| Performance Tools Makes Performance Management Possible . . . . . . . 1-3
| Performance Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
| A Recommended Approach for Setting an Objective . . . . . . . . . . . . . . 1-5
| Performance Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
| Capacity Planning—Manager Feature . . . . . . . . . . . . . . . . . . . . . . . . 1-6
| Capacity Planning—Agent Feature . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
| Performance Analysis—Manager Feature . . . . . . . . . . . . . . . . . . . . . . 1-7
| Performance Analysis—Agent Feature . . . . . . . . . . . . . . . . . . . . . . . 1-7
| Performance Analysis Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-8

Chapter 2. Starting Performance Tools . . . . . . . . . . . . . . . . . . . . . 2-1

Installing Performance Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
How Performance Tools Counts Users . . . . . . . . . . . . . . . . . . . . . . . 2-1
Printer File and Output Queues . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Start Performance Tools (STRPFRT) Command . . . . . . . . . . . . . . . . . . 2-2
Displaying the System or Job Status—Manager Feature . . . . . . . . . . . . . 2-3

Chapter 3. Collecting System Performance Data . . . . . . . . . . . . . . . 3-1

| Why Collect Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
When to Collect Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
When to End Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3
Collecting Sample or Trace Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3
Using System Defaults to Collect Data . . . . . . . . . . . . . . . . . . . . . . 3-7
Using Menus to Collect Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
Using the STRPFRMON Command to Collect Trace Data . . . . . . . . . 3-10
Collecting Performance Data Automatically . . . . . . . . . . . . . . . . . . . . 3-11
Performance Collection Setup . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13
Summary of Data Collection and Report Commands—Manager Feature . . . 3-13
System-Level Analysis—Manager Feature . . . . . . . . . . . . . . . . . . . 3-13
Job Trace Analysis—Manager Feature . . . . . . . . . . . . . . . . . . . . . 3-15
File Use and Database Structure Analysis—Manager Feature . . . . . . . 3-16

 Copyright IBM Corp. 1998 iii

 Job Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16

Chapter 4. Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1

Collecting the Right Performance Data . . . . . . . . . . . . . . . . . . . . . . . 4-2
Requesting an Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Selecting a Member . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4
Selecting Time Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4
Using a Histogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5
Analyzing Trace Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7
Using the Advisor’s Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7
Understanding Recommendations . . . . . . . . . . . . . . . . . . . . . . . . 4-8
Changing System Tuning Values . . . . . . . . . . . . . . . . . . . . . . . . 4-10
Understanding Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-11
Understanding Interval Conclusions . . . . . . . . . . . . . . . . . . . . . . 4-12
Tune System by Advisor’s Recommendations . . . . . . . . . . . . . . . . . . 4-14

Chapter 5. Displaying Performance Data . . . . . . . . . . . . . . . . . . . . 5-1

Display Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Display Performance Data by Subsystem . . . . . . . . . . . . . . . . . . . . 5-4
Display Performance Data by Job Type . . . . . . . . . . . . . . . . . . . . . 5-4
Display Performance Data by Interval . . . . . . . . . . . . . . . . . . . . . . 5-5
Display Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6
Display Job Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-7
Display Performance Data for System Resources . . . . . . . . . . . . . . . . . 5-8
Display Pool Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-8
Display Pool Interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9
Display Disk Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10
Display Disk Interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-11
Display Communications Line Detail . . . . . . . . . . . . . . . . . . . . . . 5-11
Display Remote Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-12
Display Communications Interval Data . . . . . . . . . . . . . . . . . . . . . 5-13
Display Remote Interval Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14
Display Network Interface Data . . . . . . . . . . . . . . . . . . . . . . . . . 5-15
Display Channel Interval Data . . . . . . . . . . . . . . . . . . . . . . . . . . 5-16
Display Maintenance Channel Data . . . . . . . . . . . . . . . . . . . . . . . 5-18

Chapter 6. System Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1

Work with System Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1
Single-Processor System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Multiple-Processor System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Automatic Refresh Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4
Monitoring Specific Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-5
Working with Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-6
Displaying Different Information Types . . . . . . . . . . . . . . . . . . . . . . 6-7
Accessing Work Management Functions . . . . . . . . . . . . . . . . . . . . . 6-8
Content of Database File QAITMON . . . . . . . . . . . . . . . . . . . . . . . 6-8
Print Activity Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-9
Summary Activity Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10
Detail Activity Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-12

Chapter 7. Performance Reports—Manager Feature . . . . . . . . . . . . . 7-1

A Performance Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1
Performance Report Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2
Available Performance Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3

iv Performance Tools V4R2

Which Report Do I Want? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5
Printing Performance Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5
Using Menus to Print Performance Reports . . . . . . . . . . . . . . . . . . . 7-6
Using Defaults to Print Performance Reports . . . . . . . . . . . . . . . . . 7-11
Why Performance Reports May Seem Inconsistent . . . . . . . . . . . . . 7-15
System Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17
Printing the System Report . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17
What Is the System Report? . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17
Workload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-18
Resource Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
Resource Utilization Expansion . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
Storage Pool Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
Disk Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
Communications Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
Report Selection Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
Sample System Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
Workload Section: Interactive Workload–Sample . . . . . . . . . . . . . . 7-20
Workload Section: Noninteractive Workload–Sample . . . . . . . . . . . . 7-21
Resource Utilization (First Part)–Sample . . . . . . . . . . . . . . . . . . . . 7-21
Resource Utilization (Second Part)–Sample . . . . . . . . . . . . . . . . . . 7-21
Resource Utilization Expansion (First Part)–Sample . . . . . . . . . . . . . 7-21
Resource Utilization Expansion (Second Part)–Sample . . . . . . . . . . . 7-22
Storage Pool Utilization–Sample . . . . . . . . . . . . . . . . . . . . . . . . 7-22
Disk Utilization–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22
Communications Summary–Sample . . . . . . . . . . . . . . . . . . . . . . 7-23
System Report Selection Criteria: Select Parameters–Sample . . . . . . . 7-24
System Report Selection Criteria: Omit Parameters–Sample . . . . . . . . 7-24
Report Selection Criteria: Selected Start/End Time/Date–Sample . . . . . 7-25
Report Selection Criteria: Date/Time Intervals–Sample . . . . . . . . . . . 7-25
Component Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25
Printing the Component Report . . . . . . . . . . . . . . . . . . . . . . . . . 7-25
What Is the Component Report? . . . . . . . . . . . . . . . . . . . . . . . . 7-25
Component Interval Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25
Job Workload Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26
Storage Pool Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26
Disk Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26
Input/Output Processor (IOP) Utilizations . . . . . . . . . . . . . . . . . . . 7-26
Local Work Stations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26
Remote Work Stations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-27
Exception Occurrence Summary and Interval Counts . . . . . . . . . . . . 7-27
Database Journaling Summary . . . . . . . . . . . . . . . . . . . . . . . . . 7-27
Report Selection Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-28
Sample Component Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-29
Component Interval Activity–Sample . . . . . . . . . . . . . . . . . . . . . . 7-29
Job Workload Activity–Sample . . . . . . . . . . . . . . . . . . . . . . . . . 7-29
Storage Pool Activity–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31
Disk Activity–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31
IOP Utilizations–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-32
Local Work Stations–Response Time Buckets–Sample . . . . . . . . . . . 7-33
Remote Work Stations–Response Time Buckets–Sample . . . . . . . . . . 7-33
Exception Occurrence Summary and Interval Counts–Sample . . . . . . . 7-34
Database Journaling Summary–Sample . . . . . . . . . . . . . . . . . . . . 7-34
Component Report Selection Criteria: Select Parameters–Sample . . . . 7-35
Component Report Selection Criteria: Omit Parameters–Sample . . . . . 7-36

Contents v
 Transaction Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-36
Printing the Transaction Report . . . . . . . . . . . . . . . . . . . . . . . . . 7-36
What Is the Transaction Report? . . . . . . . . . . . . . . . . . . . . . . . . 7-37
Job Summary Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-38
Interactive Throughput Section . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40
Interactive CPU Utilization Section . . . . . . . . . . . . . . . . . . . . . . . . . 7-40
Interactive Response Time Section . . . . . . . . . . . . . . . . . . . . . . . . 7-40
Scatter Diagram Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41
Interactive Program Transaction Statistics Section . . . . . . . . . . . . . . . 7-41
Seize/Lock Conflicts by Object Section . . . . . . . . . . . . . . . . . . . . . . 7-41
Special System Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41
To Print . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41
Priority-Jobtype-Pool Statistics Section . . . . . . . . . . . . . . . . . . . . . 7-42
Job Statistics Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-42
Interactive Program Statistics Section . . . . . . . . . . . . . . . . . . . . . 7-42
Individual Transaction Statistics Section . . . . . . . . . . . . . . . . . . . . 7-43
Longest Seize/Lock Conflicts Section . . . . . . . . . . . . . . . . . . . . . 7-43
Longest Holders of Seize/Lock Conflicts Section . . . . . . . . . . . . . . . 7-44
Batch Job Analysis Section . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44
Concurrent Batch Job Statistics . . . . . . . . . . . . . . . . . . . . . . . . . 7-44
Report Selection Criteria Section . . . . . . . . . . . . . . . . . . . . . . . . 7-45
Transaction Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45
To Print . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45
Job Summary Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-46
Transition Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-46
To Print . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-46
Transition Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-46
Sample Transaction Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-49
Job Summary–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-49
System Summary Data (First Part)–Sample . . . . . . . . . . . . . . . . . . 7-50
System Summary Data (Second Part)–Sample . . . . . . . . . . . . . . . . 7-51
System Summary Data (Third Part)–Sample . . . . . . . . . . . . . . . . . 7-52
Distribution of Simple, Medium, and Complex Processing Unit
Transactions–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-52
Transaction Significance–Sample . . . . . . . . . . . . . . . . . . . . . . . . 7-53
Interactive Transactions by 5-Minute Intervals–Sample . . . . . . . . . . . 7-54
Interactive Throughput by 5-Minute Intervals–Sample . . . . . . . . . . . . 7-54
Interactive CPU Utilization by 5-Minute Intervals–Sample . . . . . . . . . . 7-55
Interactive Response Time by 5-Minute Intervals–Sample . . . . . . . . . 7-55
Scatter Diagram of Interactive Transactions by 5-Minute Intervals–Sample 7-55
Interactive Program Statistics–Sample . . . . . . . . . . . . . . . . . . . . . 7-56
Summary of Seize/Lock Conflicts by Object–Sample . . . . . . . . . . . . . 7-57
Priority-Jobtype-Pool Statistics–Sample . . . . . . . . . . . . . . . . . . . . 7-58
Job Statistics–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-58
Interactive Program Statistics–Sample . . . . . . . . . . . . . . . . . . . . . 7-59
Individual Transaction Statistics–Sample . . . . . . . . . . . . . . . . . . . . 7-59
Longest Seize/Lock Conflicts–Sample . . . . . . . . . . . . . . . . . . . . . 7-59
Longest Holders of Seize/Lock Conflicts–Sample . . . . . . . . . . . . . . . 7-60
Batch Job Analysis–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-60
| Concurrent Batch Job Statistics–Sample . . . . . . . . . . . . . . . . . . . . 7-61
Report Selection Criteria-Sample . . . . . . . . . . . . . . . . . . . . . . . . 7-61
Transaction Report Option–Sample . . . . . . . . . . . . . . . . . . . . . . . 7-62
Transition Report Option–Sample . . . . . . . . . . . . . . . . . . . . . . . . 7-62
Lock Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-63

vi Performance Tools V4R2

 Printing the Lock Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-63
What Is the Lock Report? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-64
Analyzing Seize/Lock Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . 7-64
| Thread Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-65
Sample Lock Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-65
Lock Report–Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-66
Lock Report–Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-66
Job Interval Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-66
Printing the Job Interval Report . . . . . . . . . . . . . . . . . . . . . . . . . 7-66
What Is the Job Interval Report? . . . . . . . . . . . . . . . . . . . . . . . . 7-67
Interactive Job Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-67
Noninteractive Job Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 7-67
Interactive Job Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-67
Noninteractive Job Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-68
Report Selection Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-68
Sample Job Interval Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-68
Interactive Job Summary–Sample . . . . . . . . . . . . . . . . . . . . . . . 7-68
Noninteractive Job Summary–Sample . . . . . . . . . . . . . . . . . . . . . 7-69
Interactive Job Detail–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . 7-69
Noninteractive Job Detail –Sample . . . . . . . . . . . . . . . . . . . . . . . 7-70
Job Interval Report Selection Criteria: Select Parameters–Sample . . . . . 7-71
Job Interval Report Selection Criteria: Omit Parameters–Sample . . . . . 7-72
Pool Interval Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-72
Printing the Pool Interval Report . . . . . . . . . . . . . . . . . . . . . . . . 7-73
What Is the Pool Interval Report? . . . . . . . . . . . . . . . . . . . . . . . . 7-73
Subsystem Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-73
Pool Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-73
Report Selection Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-73
Sample Pool Interval Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-73
Subsystem Activity–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-74
Pool Activity–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-74
Report Selection Criteria–Sample . . . . . . . . . . . . . . . . . . . . . . . . 7-75
Resource Interval Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-76
Printing the Resource Interval Report . . . . . . . . . . . . . . . . . . . . . 7-76
What Is the Resource Interval Report? . . . . . . . . . . . . . . . . . . . . . 7-76
Disk Utilization Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-77
Disk Utilization Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-77
Communications Line Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-77
IOP Utilizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-78
Local Work Station Response Times . . . . . . . . . . . . . . . . . . . . . . 7-79
Remote Work Station Response Times . . . . . . . . . . . . . . . . . . . . 7-79
Sample Resource Interval Reports . . . . . . . . . . . . . . . . . . . . . . . . . 7-80
Disk Utilization Summary–Sample . . . . . . . . . . . . . . . . . . . . . . . 7-80
Disk Utilization Detail–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . 7-80
Communications Line Detail–SDLC Sample . . . . . . . . . . . . . . . . . . 7-81
Communications Line Detail–X.25 Sample . . . . . . . . . . . . . . . . . . 7-82
Communications Line Detail–TRLAN Sample . . . . . . . . . . . . . . . . . 7-82
Communications Line Detail–ELAN Sample . . . . . . . . . . . . . . . . . . 7-83
Communications Line Detail–DDI Sample . . . . . . . . . . . . . . . . . . . 7-84
Communications Line Detail–FRLY Sample . . . . . . . . . . . . . . . . . . 7-84
Communications Line Detail–ASYNC Sample . . . . . . . . . . . . . . . . . 7-84
Communications Line Detail–BSC Sample . . . . . . . . . . . . . . . . . . 7-85
Communications Line Detail–ISDN Network Interface Sample . . . . . . . 7-85
Communications Line Detail–NWI Maintenance Sample . . . . . . . . . . . 7-86

Contents vii
 Communications Line Detail–IDLC Samples . . . . . . . . . . . . . . . . . . 7-87
IOP Utilizations–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-87
Local Work Station Response Times–Sample . . . . . . . . . . . . . . . . . 7-88
Remote Work Station Response Times–Sample . . . . . . . . . . . . . . . 7-89
Batch Job Trace Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-89
Printing the Batch Job Trace Report . . . . . . . . . . . . . . . . . . . . . . 7-90
What Is the Batch Job Trace Report? . . . . . . . . . . . . . . . . . . . . . 7-90
Job Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-90
Job Summary Report–Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-90
Performance Trace Database Files . . . . . . . . . . . . . . . . . . . . . . . . 7-90
QTRTSUM and QTRJOBT Files . . . . . . . . . . . . . . . . . . . . . . . . 7-91
QTRJSUM File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-94
QTRDMPT File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-96
QAPTLCKD File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-101
Performance Report Columns . . . . . . . . . . . . . . . . . . . . . . . . . . 7-102

Chapter 8. Transaction Boundaries—Manager Feature . . . . . . . . . . . 8-1

Display I/O Transaction Boundary Information . . . . . . . . . . . . . . . . . . . 8-1
SNA Performance Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3
Correlation Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4
Connection Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4
Device Description Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-5
T2 Station I/O Manager Task Fields . . . . . . . . . . . . . . . . . . . . . . . 8-5
Session Traffic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6
Sending Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-7
Receiving Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-10
Intermediate Session Traffic Work Load . . . . . . . . . . . . . . . . . . . . 8-11
Control Traffic Work Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-11
Comparing Different Priority Levels . . . . . . . . . . . . . . . . . . . . . . . 8-11
APPN Control Point Performance Measurements . . . . . . . . . . . . . . . . 8-11
APPN Work Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-12
Session Setup Work Activities . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17
APPC Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-21
From System A's Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . 8-21
From System B's Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
APPC Performance Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Performance Measurement and SNADS . . . . . . . . . . . . . . . . . . . . . 8-23
SNADS Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-23
SNADS Sample Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-29
Sample Data Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-29
SNADS Performance Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-32
OS/400 File Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-33
Pass-Through Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-34
Pass-Through Performance Notes . . . . . . . . . . . . . . . . . . . . . . . 8-35
Data Queue Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-35

Chapter 9. Performance Graphics . . . . . . . . . . . . . . . . . . . . . . . . 9-1

Summary—Manager Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1
Work with Graph Formats and Packages—Manager Feature . . . . . . . . . . 9-2
Create Graph Format—Manager Feature . . . . . . . . . . . . . . . . . . . . 9-4
Graph Types—Manager Feature . . . . . . . . . . . . . . . . . . . . . . . . . 9-4
Data Types—Manager Feature . . . . . . . . . . . . . . . . . . . . . . . . . . 9-6
Legends—Manager Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-9
Create Graph Package—Manager Feature . . . . . . . . . . . . . . . . . . . 9-9

viii Performance Tools V4R2

 Change Graph Formats and Packages—Manager Feature . . . . . . . . . 9-10
Copy Graph Formats and Packages—Manager Feature . . . . . . . . . . . 9-11
Delete Graph Formats and Packages—Manager Feature . . . . . . . . . . 9-12
Display Sample Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-12
Display Package Contents—Manager Feature . . . . . . . . . . . . . . . . 9-12
Work with Historical Data—Manager Feature . . . . . . . . . . . . . . . . . . . 9-13
Create Historical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-14
Delete Historical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-15
Display Graphs and Packages—Manager Feature . . . . . . . . . . . . . . . 9-15
Display Performance Graphs—Manager Feature . . . . . . . . . . . . . . . 9-16
Display Sample Graph—Manager Feature . . . . . . . . . . . . . . . . . . . 9-17
Display Graph Package—Manager Feature . . . . . . . . . . . . . . . . . . 9-17
Select Performance Data Member—Manager Feature . . . . . . . . . . . . 9-17
Select Categories for Performance Graphs—Manager Feature . . . . . . . 9-18
Specify Graph Options—Manager Feature . . . . . . . . . . . . . . . . . . 9-18
Display Historical Graphs—Manager Feature . . . . . . . . . . . . . . . . . 9-20
Display Graph Overlay—Manager Feature . . . . . . . . . . . . . . . . . . . . 9-21

Chapter 10. Performance Utilities—Manager Feature . . . . . . . . . . . . 10-1

Job Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1
File and Process Access Group (PAG) Utilities . . . . . . . . . . . . . . . . . 10-2
Analyzing Job Flow and Transaction Performance . . . . . . . . . . . . . . . 10-2
Start Job Trace (STRJOBTRC) Command . . . . . . . . . . . . . . . . . . 10-3
End Job Trace (ENDJOBTRC) Command . . . . . . . . . . . . . . . . . . . 10-3
Print Job Trace (PRTJOBTRC) Command . . . . . . . . . . . . . . . . . . . 10-4
Analyzing the Relationship of Programs and Database Files . . . . . . . . . . 10-9
Analyze Program (ANZPGM) Command . . . . . . . . . . . . . . . . . . . . 10-9
Analyze Database File (ANZDBF) Command . . . . . . . . . . . . . . . . 10-11
Analyze Database File Keys (ANZDBFKEY) Command . . . . . . . . . . 10-13
Analyzing Process Information . . . . . . . . . . . . . . . . . . . . . . . . . . 10-16
Display Access Group (DSPACCGRP) Command . . . . . . . . . . . . . 10-16
Analyze Process Access Group (ANZACCGRP) Command . . . . . . . . 10-18

Chapter 11. Performance Explorer . . . . . . . . . . . . . . . . . . . . . . . 11-1

Do I Need Performance Explorer? . . . . . . . . . . . . . . . . . . . . . . . . . 11-1
Who Needs Performance Explorer . . . . . . . . . . . . . . . . . . . . . . . 11-1
When You Need Performance Explorer . . . . . . . . . . . . . . . . . . . . 11-2
Comparison of Explorer to Other Performance Tools . . . . . . . . . . . . . . 11-2
Performance Explorer and Advisor Functions . . . . . . . . . . . . . . . . . 11-2
Performance Explorer and Performance Monitor . . . . . . . . . . . . . . . 11-2
Benefits of Performance Explorer . . . . . . . . . . . . . . . . . . . . . . . . . 11-3
How Performance Explorer Works . . . . . . . . . . . . . . . . . . . . . . . . . 11-3
Performance Explorer Definitions . . . . . . . . . . . . . . . . . . . . . . . . . 11-4
General Flow of the Performance Explorer . . . . . . . . . . . . . . . . . . 11-5
Creating a Performance Explorer Definition . . . . . . . . . . . . . . . . . . . . 11-6
Starting the Performance Explorer . . . . . . . . . . . . . . . . . . . . . . . . . 11-7
Ending the Performance Explorer . . . . . . . . . . . . . . . . . . . . . . . . . 11-7
Deleting Performance Explorer Data . . . . . . . . . . . . . . . . . . . . . . . . 11-8
| Creating and Printing Performance Explorer Reports . . . . . . . . . . . . . . 11-8
Finding Your Performance Explorer Definitions . . . . . . . . . . . . . . . . . 11-9
Types of Performance Explorer Reports . . . . . . . . . . . . . . . . . . . . . 11-9
Common Report Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-9
Report-Specific Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-15
Mapping OPM High-Level Language (HLL) Statements to Source Code . . 11-35

Contents ix
 Chapter 12. Managing the Performance Tools Configuration . . . . . . . 12-1
Work with Functional Areas—Manager Feature . . . . . . . . . . . . . . . . . 12-2
Creating a Functional Area—Manager Feature . . . . . . . . . . . . . . . . 12-3
Changing a Functional Area—Manager Feature . . . . . . . . . . . . . . . 12-4
Deleting a Functional Area—Manager Feature . . . . . . . . . . . . . . . . 12-5
Copying a Functional Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-5
Delete Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-5
Copy Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-7
Convert Performance Data (CVTPFRDTA) Command . . . . . . . . . . . . . 12-8

Chapter 13. A Problem Analysis Case Study . . . . . . . . . . . . . . . . . 13-1

Introduction to Performance Analysis . . . . . . . . . . . . . . . . . . . . . . . 13-1
The Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-2
The Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-2
The Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-3
The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-3
Checking the System’s Performance . . . . . . . . . . . . . . . . . . . . . . 13-4
Reviewing the End-User Survey Results . . . . . . . . . . . . . . . . . . . 13-10
Analyzing System Performance . . . . . . . . . . . . . . . . . . . . . . . . 13-11
Understanding the Symptoms of the Problem . . . . . . . . . . . . . . . . 13-21
Analyzing the Data—Manager Feature . . . . . . . . . . . . . . . . . . . . 13-23
Case Study Data Reports—Manager Feature . . . . . . . . . . . . . . . . 13-28
Finding the Cause and Correcting the Problem . . . . . . . . . . . . . . . 13-32
Final Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-33

Chapter 14. Working with Historical Data—Agent Feature . . . . . . . . 14-1

Create Historical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-2
Delete Historical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-3

Chapter 15. Managing the Performance Data—Agent Feature . . . . . . 15-1

Delete Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1
Copy Performance Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-2
Convert Performance Data (CVTPFRDTA) Command . . . . . . . . . . . . . 15-4
| Convert Performance Thread Data (CVTPFRTHD) Command . . . . . . . . . 15-6

Appendix A. Performance Tools CL Commands . . . . . . . . . . . . . . A-1

ADDPEXDFN (Add Performance Explorer Definition) Command . . . . . . . A-2
ANZACCGRP (Analyze Access Group) Command . . . . . . . . . . . . . . . A-5
ANZDBF (Analyze Database File) Command . . . . . . . . . . . . . . . . . . . A-6
ANZDBFKEY (Analyze Database File Keys) Command . . . . . . . . . . . . A-7
ANZPFRDTA (Analyze Performance Data) Command . . . . . . . . . . . . . A-7
ANZPGM (Analyze Program) Command . . . . . . . . . . . . . . . . . . . . . A-8
CHGFCNARA (Change Functional Area) Command . . . . . . . . . . . . . . A-9
CHGGPHFMT (Change Graph Format) Command . . . . . . . . . . . . . . . A-10
CHGGPHPKG (Change Graph Package) Command . . . . . . . . . . . . . . A-13
CHGJOBTYP (Change Job Type) Command . . . . . . . . . . . . . . . . . . . A-14
CHGPEXDFN (Change Performance Explorer Definition) Command . . . . . A-14
CPYFCNARA (Copy Functional Area) Command . . . . . . . . . . . . . . . . A-17
CPYGPHFMT (Copy Graph Format) Command . . . . . . . . . . . . . . . . . A-17
CPYGPHPKG (Copy Graph Package) Command . . . . . . . . . . . . . . . . A-18
CPYPFRDTA (Copy Performance Data) Command . . . . . . . . . . . . . . . A-19
CRTFCNARA (Create Functional Area) Command . . . . . . . . . . . . . . . A-20
CRTGPHFMT (Create Graph Format) Command . . . . . . . . . . . . . . . . A-21
CRTGPHPKG (Create Graph Package) Command . . . . . . . . . . . . . . . A-24

x Performance Tools V4R2

 CRTHSTDTA (Create Historical Data) Command . . . . . . . . . . . . . . . . A-25
DLTFCNARA (Delete Functional Area) Command . . . . . . . . . . . . . . . . A-26
DLTGPHFMT (Delete Graph Format) Command . . . . . . . . . . . . . . . . . A-27
DLTGPHPKG (Delete Graph Package) Command . . . . . . . . . . . . . . . A-27
DLTHSTDTA (Delete Historical Data) Command . . . . . . . . . . . . . . . . A-28
DLTPEXDTA (Delete Performance Explorer Data) Command . . . . . . . . . A-29
DLTPFRDTA (Delete Performance Data) . . . . . . . . . . . . . . . . . . . . . A-29
DSPACCGRP (Display Access Group) Command . . . . . . . . . . . . . . . . A-30
DSPHSTGPH (Display Historical Graph) Command . . . . . . . . . . . . . . . A-32
DSPPFRDTA (Display Performance Data) Command . . . . . . . . . . . . . . A-35
DSPPFRGPH (Display Performance Graph) Command . . . . . . . . . . . . A-36
ENDJOBTRC (End Job Trace) Command . . . . . . . . . . . . . . . . . . . . A-41
ENDPEX (End Performance Explorer) Command . . . . . . . . . . . . . . . . A-42
PRTACTRPT (Print Activity Report) Command . . . . . . . . . . . . . . . . . A-43
PRTCPTRPT (Print Component Report) Command . . . . . . . . . . . . . . . A-45
PRTJOBRPT (Print Job Report) Command . . . . . . . . . . . . . . . . . . . . A-48
PRTJOBTRC (Print Job Trace) Command . . . . . . . . . . . . . . . . . . . . A-51
PRTLCKRPT (Print Lock Report) Command . . . . . . . . . . . . . . . . . . . A-52
PRTPEXRPT (Print Performance Explorer Report) Command . . . . . . . . . A-53
PRTPOLRPT (Print Pool Report) Command . . . . . . . . . . . . . . . . . . . A-56
PRTRSCRPT (Print Resource Report) Command . . . . . . . . . . . . . . . . A-59
PRTSYSRPT (Print System Report) Command . . . . . . . . . . . . . . . . . A-60
PRTTNSRPT (Print Transaction Report) Command . . . . . . . . . . . . . . . A-64
PRTTRCRPT (Print Trace Report) Command . . . . . . . . . . . . . . . . . . A-66
RMVPEXDFN (Remove Performance Explorer Definition) Command . . . . . A-67
STRJOBTRC (Start Job Trace) Command . . . . . . . . . . . . . . . . . . . . A-68
STRPEX (Start Performance Explorer) Command . . . . . . . . . . . . . . . . A-69
STRPFRG (Start Performance Graphics) Command . . . . . . . . . . . . . . A-70
STRPFRT (Start Performance Tools) Command . . . . . . . . . . . . . . . . . A-70
WRKFCNARA (Work with Functional Areas) Command . . . . . . . . . . . . A-71
WRKSYSACT (Work with System Activity) Command . . . . . . . . . . . . . A-72

Appendix B. Defining Transaction Boundaries . . . . . . . . . . . . . . . B-1

Elements of Response Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-1
Differences in the Transaction Response Reports . . . . . . . . . . . . . . . . B-2
Operational Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . B-4

Appendix C. Correlation of the System/36 and AS/400 System

Performance Parameters—Manager Feature . . . . . . . . . . . . . . . . . C-1

Appendix D. Comparison of Performance Tools . . . . . . . . . . . . . . D-1

Comparison of Functions, Menu Options, and Commands . . . . . . . . . . . D-1

Appendix E. Managing AS/400 System Performance in a Network . . . E-1

Planning the Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-1
Implementing the Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-1
Additional Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-4

Appendix F. Performance Checklist—Manager Feature . . . . . . . . . . . F-1

Planning for Performance and Tuning . . . . . . . . . . . . . . . . . . . . . . . . F-1
Basic Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F-2
| Work with System Status Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . F-2
| Work with Disk Status Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . F-3
| Work with System Activity Tips . . . . . . . . . . . . . . . . . . . . . . . . . . F-3

Contents xi
 General Tuning Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F-4
General Performance Facts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F-4

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-1

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-3

xii Performance Tools V4R2

Notices
This information was developed for products and services offered in the U.S.A. IBM
may not offer the products, services, or features discussed in this document in
other countries. Consult your local IBM representative for information on the pro-
ducts and services currently available in your area. Any reference to an IBM
product, program, or service is not intended to state or imply that only that IBM
product, program, or service may be used. Any functionally equivalent product,
program, or service that does not infringe any IBM intellectual property right may be
used instead. However, it is the user's responsibility to evaluate and verify the oper-
ation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter
described in this document. The furnishing of this document does not give you any
license to these patents. You can send license inquiries, in writing, to:

IBM Director of Licensing

IBM Corporation
500 Columbus Avenue
Thornwood, NY 10594
U.S.A.

For license inquiries regarding double-byte (DBCS) information, contact the IBM
Intellectual Property Department in your country or send inquiries, in writing, to:

IBM World Trade Asia Corporation

Licensing
2-31 Roppongi 3-chome, Minato-ku
Tokyo 106, Japan

The following paragraph does not apply to the United Kingdom or any other
country where such provisions are inconsistent with local law: INTERNA-
TIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS PUBLICATION
“AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR
PURPOSE. Some states do not allow disclaimer of express or implied warranties in
certain transactions, therefore, this statement may not apply to you.

This information could include technical inaccuracies or typographical errors.

Changes are periodically made to the information herein; these changes will be
incorporated in new editions of the publication. IBM may make improvements
and/or changes in the product(s) and/or the program(s) described in this publication
at any time without notice.

Licensees of this program who wish to have information about it for the purpose of
enabling: (i) the exchange of information between independently created programs
and other programs (including this one) and (ii) the mutual use of the information
which has been exchanged, should contact:

 Copyright IBM Corp. 1998 xiii

 IBM Corporation
Software Interoperability Coordinator
3605 Highway 52 N
Rochester, MN 55901-7829
U.S.A.

Such information may be available, subject to appropriate terms and conditions,

including in some cases, payment of a fee.

The licensed program described in this information and all licensed material avail-
able for it are provided by IBM under terms of the IBM Customer Agreement or any
equivalent agreement between us.

Any performance data contained herein was determined in a controlled environ-

ment. Therefore, the results obtained in other operating environments may vary sig-
nificantly. Some measurements may have been made on development-level
systems and there is no guarantee that these measurements will be the same on
generally available systems. Furthermore, some measurement may have been esti-
mated through extrapolation. Actual results may vary. Users of this document
should verify the applicable data for their specific environment.

This information contains examples of data and reports used in daily business oper-
ations. To illustrate them as completely as possible, the examples include the
names of individuals, companies, brands, and products. All of these names are
fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.

COPYRIGHT LICENSE:

This information contains sample application programs in source language, which

illustrates programming techniques on various operating platforms. You may copy,
modify, and distribute these sample programs in any form without payment to IBM,
for the purposes of developing, using, marketing or distributing application pro-
grams conforming to the application programming interface for the operating plat-
form for which the sample programs are written. These examples have not been
thoroughly tested under all conditions. IBM, therefore, cannot guarantee or imply
reliability, serviceability, or function of these programs. You may copy, modify, and
distribute these sample programs in any form without payment to IBM for the pur-
poses of developing, using, marketing, or distributing application programs con-
forming to IBM's application programming interfaces.

If you are viewing this information softcopy, the photographs and color illustrations
may not appear.

Programming Interface Information

This publication is intended to help you to achieve high system performance. This
publication documents General-Use Programming Interface and Associated Guid-
ance Information provided by IBM Performance Tools for AS/400 (5769-PT1).

General-Use programming interfaces allow the customer to write programs that

obtain the services of IBM Performance Tools for AS/400 and IBM Operating
System/400 licensed programs.

xiv Performance Tools V4R2

 This publication documents Product-Sensitive Programming Interface and Associ-
ated Guidance Information.

Product-Sensitive programming interfaces allow the customer installation to perform

tasks such as diagnosing, modifying, monitoring, repairing, tailoring, or tuning of
this IBM software product. Use of such interfaces creates dependencies on the
detailed design or implementation of the IBM software product. Product-Sensitive
programming interfaces should be used only for these specialized purposes.
Because of their dependencies on detailed design and implementation, it is to be
expected that programs written to such interfaces may need to be changed in order
to run with new product releases or versions, or as a result of service.

Trademarks
The following terms are trademarks of the IBM Corporation in the United States or
other countries or both:

Advanced Function Printing

Advanced 36
AFP
Application System/400
APPN
AS/400
AS/400e
C/400
Client Access
IBM
Information Assistant
Intelligent Printer Data Stream
IPDS
MVS
OfficeVision/400
Operating System/400
OS/400
Personal Computer AT
Personal System/2
RPG/400
System/36
SystemView
400

Microsoft, Windows, Windows NT, and the Windows 95 logo are registered trade-
marks of Microsoft Corporation.

UNIX is a registered trademark in the United States and other countries licensed
exclusively through X/Open Company Limited

Java and HotJava are trademarks of Sun Microsystems, Inc.

Other company, product, and service names may be trademarks or service marks
of others.

Notices xv
xvi Performance Tools V4R2
About Performance Tools (SC41-5340)
This book explains how to use performance tools to collect data about the perform-
ance of a system, job, or program. It also explains how to analyze and print the
data to help identify and correct any problems.

The book addresses both the Manager feature and the Agent feature. Most
sections are marked to indicate the feature to which the information applies. If a
section is not marked as Manager feature or Agent feature, the section applies to
both.

Who should read this book

This book is intended for anyone who has to perform data collections and analyze
performance data.

In this book, personal computer means an IBM Personal Computer such as a 5170
Personal Computer AT or an 8560 Personal System/2.

The performance estimates presented are approximations which are believed to be

sound. The degree of success that you may achieve in the use of IBM equipment
and programs is dependent upon a number of factors, many of which are not under
IBM's control. Thus, IBM neither warrants nor guarantees that you can or will
achieve similar results. It is your responsibility to validate the estimates furnished
and to determine their relevance to your operation.

Any configuration recommended by the capacity planner of the Manager feature

should be verified with your marketing representative because the capacity planner
does not consider all attachable devices.

AS/400 Operations Navigator

AS/400 Operations Navigator is a powerful graphical interface for Windows 95/NT
clients. With AS/400 Operations Navigator, you can use your Windows 95/NT skills
to manage and administer your AS/400 systems. You can work with database
administration, file systems, Internet network administration, users, and user
groups. You can even schedule regular system backups and display your hardware
and software inventory. Figure 0-1 on page xviii shows an example of the display.

 Copyright IBM Corp. 1998 xvii

 Figure 0-1. AS/400 Operations Navigator Display

IBM recommends that you use this new interface. It is simple to use and has great
online information to guide you.

You can access the AS/400 Operations Navigator from the Client Access folder by
double-clicking the AS/400 Operations Navigator icon. You can also drag this icon
to your desktop for even quicker access.

While we develop this interface, you will still need to use the familiar AS/400 “green
screens” to do some of your tasks. You can find information to help you in this
book and online.

Prerequisite and related information

You should be familiar with the information about performance analysis as
described in the Work Management book before using this book.

The menus and displays shown in this book are used by the Manager feature.
Displayes used by the Agent feature may contain fewer options than those shown
for the Manager feature.

For information about Advanced 36 publications, see the Advanced 36 Information

Directory, SC21-8292, in the AS/400 Softcopy Library.

| For information about other AS/400 publications (except Advanced 36), see either
| of the following:
| Ÿ The Publications Reference, SC41-5003, in the AS/400 Softcopy Library.
| Ÿ The AS/400 online library is available on the World Wide Web at the following
| uniform resource locator (URL) address:
| http://as4ððbks.rochester.ibm.com/

For a list of related publications, see the “Bibliography” on page X-1.

xviii Performance Tools V4R2

| Information available on the World Wide Web
| In addition to the AS/400 online library on the World Wide Web, you can access
| other information from the AS/400 Technical Studio at the following URL address:
| http://www.as4ðð.ibm.com/techstudio

| How to send your comments

| Your feedback is important in helping to provide the most accurate and high-quality
| information. If you have any comments about this book or any other AS/400 doc-
| umentation, fill out the readers' comment form at the back of this book.
| Ÿ If you prefer to send comments by mail, use the readers' comment form with
| the address that is printed on the back. If you are mailing a readers' comment
| form from a country other than the United States, you can give the form to the
| local IBM branch office or IBM representative for postage-paid mailing.
| Ÿ If you prefer to send comments by FAX, use either of the following numbers:
| – United States and Canada: 1-800-937-3430
| – Other countries: 1-507-253-5192
| Ÿ If you prefer to send comments electronically, use this network ID:
| – IBMMAIL, to IBMMAIL(USIB56RZ)
| – [email protected]
| Be sure to include the following:
| Ÿ The name of the book.
| Ÿ The publication number of the book.
| Ÿ The page number or topic to which your comment applies.

About Performance Tools (SC41-5340) xix

xx Performance Tools V4R2
| Summary of Changes
| Changes to this book for Version 4 Release 2 include the following:
| Ÿ Removal of Chapter 3, Performance Tuning. This chapter was a duplicate of
| Chapter 14 in the Work Management book. The Performance Tuning chapter
| remains in the Work Management book.
| Ÿ Removal of the parameter descriptions for the CL commands. The syntax dia-
| grams, purpose description, and examples remain in the appendix.
| Ÿ You can stop the automatic refresh function on the Work with System Activity
| display by pressing a function key.
| Ÿ A brief description of the Convert Performance Thread Data (CVTPFRTHD)
| command. The details are in the Work Management book.
| Ÿ The *LGLDBIO value was added to the Y axis (YAXIS) parameter on the
| Create Graph Format (CRTGPHFMT) command and the Change Graph Format
| (CHGGPHFMT) command. This value displays the logical I/Os for a particular
| job.
| Ÿ The concept of what constitutes a job is changed with the introduction of thread
| information. A job contains one or more threads, one initial thread and sec-
| ondary threads. The Work with System Activity display shows thread informa-
| tion.
| Ÿ Changes to Reports:
| – Thread information is reflected in many of the reports when the collections
| contain thread information. Some reports add the thread identifier to the
| user name column when the report shows a secondary thread. The user
| name in this case is the user name portion of the qualified job name that is
| running the thread. Other reports show a separate thread identifier column.
| – The line speeds for the Ethernet adapter and the token-ring adapter show
| as half duplex or full duplex.
| – The line speed for asynchronous transfer mode (ATM) shows as half
| duplex or full duplex.
| – The Batch Thread Analysis section of the Transaction Report was renamed
| to Concurrent Batch Job Statistics.
| – Disk CPU utilization is reported under the Dsk CPU Util column.
| Ÿ Performance explorer improves the filtering capabilities for collected trace data.
| Ÿ Performance explorer provides an OUTFILE parameter on the Print Perform-
| ance Explorer Report command. The OUTFILE parameter allows you to write
| queries against the trace data.
| Ÿ Performance explorer provides a new report format, *BASIC. The *BASIC
| Report provides summary information that includes the definition, run, and task
| information sections for any of the other report types.

 Copyright IBM Corp. 1998 xxi

xxii Performance Tools V4R2
| Chapter 1. Introduction to Performance Tools
| How much do you invest in managing the performance of your system? The needs
| of your business change, sometimes sooner than you expect. To respond to busi-
| ness changes effectively, your system must change too. Managing your system, at
| first glance, might seem like just another time-consuming job. But the systems man-
| agement investment pays off soon because the system runs more efficiently and
| this is reflected in your business. It is efficient because changes are planned and
| managed.

| Using Performance Tools helps you gain insight into the many built-in performance
| management features already working for you in OS/400. These features include
| dynamic tuning, expert cache, job priorities, activity levels, and pool sizes. You can
| also identify ways to use these services better. You might find specific actions for
| your system that the “built-in” OS/400 features do not address.

| Why Manage Performance?

| Does your system perform as well as it could? If it can handle today's jobs, is that
| enough? What about the future—are you prepared for growth?

| Sometimes good performance just happens. In those cases, the system has plenty
| of resources to get the job done. But there will be times when those resources are
| not in the right place. Maybe you have added systems and clients to the network,
| or the production has increased and the workload is significantly changed. Or, more
| often, workload changes in small, nearly invisible increments and one day perform-
| ance just is not as good anymore. That is why you have to plan ahead for your
| system to be at peak performance, especially in a quick-paced business.

| As a result, it is important to manage performance effectively for both the long term
| and the short term. In the short term, understanding the performance components
| of your system helps you react quickly when a performance problem occurs at a
| crucial time. It may also allow you to defer upgrading for a few months. In the long
| term, if you plan for a more efficient system, you prevent potential performance
| problems from developing. You also ensure that you have enough capacity on the
| system to handle your workloads. In addition, your users get the service they
| expect. Maintaining good performance requires that you understand, plan, and
| manage performance.

| Effectively managing performance really comes in the form of your own perform-
| ance plan. Performance management is necessary to optimize the use of a system
| and its associated services, such as the efficiency of communications lines. Per-
| formance management is a strategy for planning, implementing, controlling, and
| measuring computer-based tasks to achieve performance that is acceptable. But
| the concept of acceptable performance is relative to where your business started
| and where it is going, as well as the users and their needs.

 Copyright IBM Corp. 1998 1-1

| Performance Capabilities—OS/400 and Performance Tools
| Managing performance includes tasks like collecting data, monitoring, and tuning.
| OS/400 provides these basic functions. Performance Tools builds on these basic
| components to help you more thoroughly assess current performance needs and
| plan for future capacity needs.

| OS/400 Has Performance Management Basics

| Before understanding Performance Tools, you should thoroughly understand the
| performance components of OS/400 and have a good understanding of work man-
| agement. OS/400 can do basic collecting, monitoring, and tuning.

| For Collecting
| The OS/400 performance monitor collects the data as a batch job that is run by the
| operating system.

| For Monitoring
| OS/400 provides three commands that display or print performance-related
| information:
| Ÿ Work with Active Jobs (WRKACTJOB)
| Ÿ Work with Disk Status (WRKDSKSTS)
| Ÿ Work with System Status (WRKSYSSTS)

| For Tuning
| Did you know that the system can tune itself dynamically? A system value, perform-
| ance adjust (QPFRADJ), allows you to have the system tuned dynamically. Per-
| formance tuning is the ability to adjust some components on the system to better
| manage your workloads, manually or automatically.

| The expert cache function of OS/400 is the method of tuning to automatically adjust
| the storage pool paging. Also, the QDYNPTYSCD (dynamic priority scheduling)
| system value allows you to turn on and off the dynamic priority scheduler.

| Examples of adjustments include the following:

| Ÿ Changing system shared pools such as *MACHINE, *BASE, or *INTERACT
| Ÿ Changing a pool's activity level

| Performance Tools Builds on OS/400 Function

| Performance Tools builds on the basic collecting, monitoring, and tuning capabilities
| of OS/400. For example, you can collect performance data by using the perform-
| ance monitor; then Performance Tools provides the capability of:
| Ÿ Analyzing performance data
| Ÿ Summarizing it into reports
| Ÿ Creating graphs to show trends
| Ÿ Advising how to adjust the system components to achieve better performance

| Also, although the OS/400 commands display data on a real-time basis, they are
| tedious. They require repetitive use for an operator who also needs to determine

1-2 Performance Tools V4R2

| the problem areas and can only make a limited set of changes. The Work with
| System Activity (WRKSYSACT) command can automatically refresh and use less
| system resource than the Work with Active Jobs (WRKACTJOB) command. To
| determine the causes of problems you find, you need Performance Tools to effec-
| tively and quickly analyze the data.

| Finally, the tuning function you have through OS/400 is minimal in detail when com-
| pared to Performance Tools. Often small adjustments can be made. However, if
| you have a complex system network, the problems with performance are often not
| isolated enough to solve through adjusting storage pools and activity levels. You
| need a more thorough analysis of performance data to find the cause of the prob-
| lems and optimize performance.

| Manager Features and Agent Features

| Similar to the way that the Performance Tools product builds on the base of
| OS/400, the Performance Tools Manager feature builds on the Agent feature. The
| Performance Tools Manager feature is a full-function package, intended to be used
| on the central site system in a distributed environment or on a single system.

| The Performance Tools Agent feature, with a subset of the Manager function, is a
| lower-priced package with the more basic functions. In a distributed environment,
| the Agent feature works well for managed systems in the network because the data
| can be sent to the Manager if detailed analysis is required. It is also an effective
| tool for sites that need a reasonable level of self-sufficiency but have no expert
| skills available.

| Performance Tools Makes Performance Management Possible

| Performance Tools makes this planning possible. More specifically, they are tools
| that:
| Ÿ Predict performance that is based on your workloads and configurations and
| models of your growth.
| Ÿ Collect, access, and work with performance data. The performance data avail-
| able through these tools includes real-time graphical and character-based dis-
| plays, historical data, charted graphs, interactive displays, printed reports, and
| detailed trace data. The various forms make several types of analysis possible.
| Ÿ Use an expert system that assesses the situation and offers advice to optimize
| performance.
| Ÿ Track system use and growth at each location by using long-term historical
| graphs.
| Ÿ Identify applications and system resources that are causing problems with
| system performance.

| Performance Objectives
| Before you use Performance Tools, however, you must establish performance
| objectives for:
| Ÿ Throughput and response time for interactive jobs
| Ÿ Throughput for batch jobs
| Ÿ Resource utilizations for the system

Chapter 1. Introduction to Performance Tools 1-3

| Figure 1-1. Performance Tools Helps You Achieve Peak Performance

| You should realize that by concentrating on one objective, you could adversely
| affect another. For example, if your users want fast response time, you need to
| design and operate your system so that your users receive stable response time
| over a range of system loads. This choice, however, could cause batch jobs to run
| slower.
| Note: Two types of response time are discussed in this guide. Internal response
| time is the AS/400 system host response time. External response time is
| the end-user response time and includes communications time for both
| locally and remotely attached display stations.

| When response time is mentioned, unless stated otherwise, assume that it is

| internal response time.

| For additional information, see “Elements of Response Time” on page B-1.

1-4 Performance Tools V4R2

| After you set the performance objectives for interactive and batch jobs, install and
| use Performance Tools. The primary functions of Performance Tools, as described
| in the following sections, are related. Use each function to make the greatest use of
| your system’s performance.

| A Recommended Approach for Setting an Objective

| Managing performance means performing the following items to monitor your
| system. Once you have collected and analyzed the data, the changes you make
| should be small. If you make slight changes, you can determine the affect of each
| specific change. Also, you should prioritize the changes. Continue making changes
| and measuring performance until you reach your objectives.
| 1. Set performance goals
| Ÿ Set goals that match the demands of your business
| Ÿ Identify areas of the system where an improvement in performance can
| make an impact on your business
| Ÿ Make the goals reasonable and ones that can be measured
| 2. Collect performance data
| Ÿ Record performance measurement daily or weekly
| – Always save performance measurement data before installing a
| new software release, a major hardware upgrade, a new applica-
| tion, or a large number of additional workstations or jobs.
| – Each measurement period should include "typical" medium to heavy
| workloads.
| – A good starting measurement period is two hours with 15 minutes
| samples. You should use separate measurement periods if there are
| significantly different workloads or job types that are based on time of
| day. If you are trying to solve a problem, 5-minute intervals are better
| for “catching” the problem.
| Ÿ Check and analyze performance data
| – Summarize the collected data
| - Compare the data to objectives or resource guidelines.
| – Perform monthly trend analysis that includes at least the previous three
| months of summarized data
| - As time progresses, include at least six months of summary data to
| ensure that a trend is consistent. Make decisions based on at least
| three months of trend information and your knowledge of upcoming
| system demands.
| Ÿ Tune performance
| – Tune performance whenever guidelines are not met.
| – Analyze performance data to catch situations before they become prob-
| lems.
| - Performance data indicates objectives that have not been met
| - Trend analysis indicates if resource consumption is increasing sig-
| nificantly or performance objectives are at or past guideline values

Chapter 1. Introduction to Performance Tools 1-5

| – Plan capacity when:
| - Trend analysis shows a significant growth in resource utilization
| - A major new application or set of additional jobs will be added to
| the current hardware configuration. This could include additional
| interactive workstations or new batch jobs.
| - Business plans are reviewed, and change is expected

| Performance Measurement
| When you first start to use Performance Tools, use the Start Performance Tools
| (STRPFRT) command to show the IBM Performance Tools menu. You should
| collect performance data (by using the Start Performance Monitor (STRPFRMON)
| command). When you use the Manager feature, you can produce a system report
| so you have a measure of how your system currently operates. Use the Print
| System Report (PRTSYSRPT) command. “Using Defaults to Print Performance
| Reports” on page 7-11 describes this process. When you use the Manager feature,
| you can collect performance data automatically by using either the Work with Per-
| formance Collection (WRKPFRCOL) command or the job scheduling functions.

| Collecting and analyzing performance data regularly determines if you meet your
| performance objectives. In this way, you accumulate a history of system perform-
| ance. This history is important in managing system performance, especially if your
| environment is one of growth and change. By monitoring system performance, you
| may avoid excessive use of your system’s resources.

| For more information, see Chapter 3, Collecting System Performance Data.

| Capacity Planning—Manager Feature

| To estimate your system resource utilization as your workload or environment
| grows, use the capabilities of BEST/1** for the AS/400 system. BEST/1 is the
| capacity planning tool for the AS/400 system. Following BEST/1 recommendations
| will help you maintain satisfactory system performance and system resource utiliza-
| tions.

| Do capacity planning before you make changes, such as adding new applications
| or altering the system configuration. See the BEST/1 Capacity Planning Tool book
| for more information.

| Capacity Planning—Agent Feature

| The Agent feature provides the ability to create BEST/1 models from performance
| data. These models can be analyzed by using the BEST/1 support in the Manager
| feature.

| See the BEST/1 Capacity Planning Tool book for more information.

1-6 Performance Tools V4R2

| Performance Analysis—Manager Feature
| After you review the performance measurements, you might want to see more
| detailed performance data. Use the Print System Report (PRTSYSRPT) and Print
| Component Report (PRTCPTRPT) commands to help you decide if further analysis
| is necessary. Chapter 7, Performance Reports—Manager Feature shows examples
| of these reports.

| To provide more detail, you can also produce reports that use trace data (specify
| TRACE(*ALL) on the STRPFRMON command). Use the Print Transaction Report
| (PRTTNSRPT) command to help you do further analysis of performance problems
| you may be experiencing.

| The advisor, that is described in Chapter 4, Advisor, analyzes collected perform-

| ance data and produces conclusions and recommendations for improving system
| performance. You can have the advisor put the recommendations into effect. You
| can use the conclusions and recommendations to help you decide how to adjust
| system tuning values.

| See Chapter 10, Performance Utilities—Manager Feature, for an explanation and

| examples of other utilities you can use to analyze the performance of applications
| on your system. See Chapter 5, Displaying Performance Data, for an explanation
| on interactively displaying performance data.

| The performance explorer is a tool that finds the causes of performance problems
| that cannot be identified by using tools that do general performance monitoring.
| Chapter 11, Performance Explorer describes the performance explorer.

| “Summary of Data Collection and Report Commands—Manager Feature” on

| page 3-13 provides a summary of data collection commands and reporting com-
| mands.

| Performance Analysis—Agent Feature

| The advisor, that is described in Chapter 4, Advisor , analyzes collected perform-
| ance data and produces conclusions and recommendations for improving system
| performance. You can have the advisor put the recommendations into effect. You
| can use the conclusions and recommendations to help you decide how to adjust
| system tuning values.

| See Chapter 5, Displaying Performance Data, for an explanation on interactively

| displaying performance data.

| The performance explorer is a tool that finds the causes of performance problems
| that cannot be identified by using tools that do general performance monitoring.
| Chapter 11, Performance Explorer describes the performance explorer.

| See Chapter 14, Working with Historical Data—Agent Feature, for an explanation
| on how to use the option to create historical data from performance data. The his-
| torical data will help show the trends in your system performance.

| See Appendix E, Managing AS/400 System Performance in a Network, for an

| example of automating performance data collection and analysis.

Chapter 1. Introduction to Performance Tools 1-7

| Performance Analysis Overview
| Performance analysis is a method for investigating, measuring, and correcting defi-
| ciencies so that system performance meets the user’s expectations. It does not
| matter much that the system is a computer; it could be an automobile or a washing
| machine. The problem-solving approach is essentially the same:
| 1. Understand the symptoms of the problem.
| 2. Use tools to measure and define the problem.
| 3. Isolate the cause.
| 4. Correct the problem.
| 5. Use tools to verify the correction.

| Initially, the analyst knows that the user is not satisfied with the way the system is
| working. It is running “too slow,” “too noisy,” or “too hot,” and so on. The analyst,
| mechanic, or repair person must first understand what the problem really is. The
| best way to find out is to observe the problem condition firsthand. Can you confirm
| the user’s complaint? If you cannot, get as much information as possible from users
| who have experienced the problem. Look and listen for the problem descriptions
| most in common among the users.

| The key to success with any performance issue is to have a clear definition of the
| users’ performance criteria. In other words, given the application mix, what do users
| want from the system in terms of interactive response time, batch throughput, and
| processing requirements? For example, a system that supports an interactive order
| entry application may have a response time criterion to ensure that customers do
| not perceive abnormal delays. Another criterion may require that end-of-day proc-
| essing be completed by a specific time. With these requirements, you can establish
| performance objectives around system resource utilization guidelines. With a clear
| statement of goals and objectives, performance analysis can proceed on a firm
| basis.

| Once you understood the objectives, it is important to assess whether the hardware
| configuration is adequate to support the workload. Is there enough CPU capacity?
| Is the main storage sufficient for the application mix? Answering these questions
| first, perhaps through capacity modeling techniques, prevents needless effort later.

| With an understanding of the symptoms of the problem and the objectives to be

| met, the analyst can formulate a hypothesis that may explain the cause of the
| problem. The analyst can use certain Operating System/400 (OS/400) commands
| and the Performance Tools to measure the system performance.

| Reviewing the measured data helps to further define the problem and helps to vali-
| date or reject the hypothesis. Once the apparent cause or causes have been iso-
| lated, a solution can be proposed. When you handle one solution at a time, you
| can re-design and test programs. Again, the analyst’s tools can, in many cases,
| measure the effectiveness of the solution and look for possible side-effects.

| To achieve optimum performance, one must recognize the interrelationship among

| the critical system resources and attempt to balance these resources, namely CPU,
| disk, main storage, and, for communications, remote lines. Each of these resources
| may cause a performance degradation.

1-8 Performance Tools V4R2

| Improvements to system performance, whether to interactive throughput, interactive
| response time, batch throughput, or some combination, may take many forms, from
| simply adjusting activity level or pool size to changing the application code itself. In
| this instance, an activity level is a characteristic of a subsystem that specifies the
| maximum number of jobs that can compete at the same time for the processing
| unit.

| Ultimately, however, any improvement can come only through analysis of the crit-
| ical resources (CPU, main storage, disk, and remote lines) and contention for
| system and application objects.

Chapter 1. Introduction to Performance Tools 1-9

1-10 Performance Tools V4R2
Chapter 2. Starting Performance Tools
This chapter explains how to install and set up your Performance Tools. Informa-
tion on how to use the Start Performance Tools (STRPFRT) command is also pro-
vided.

Installing Performance Tools

To install the Performance Tools product, you need a user profile with save system
(*SAVSYS) authority. You can use the system operator profile to obtain this
authority.

Performance Tools must run in a library named QPFR. If a library by this name is
on your system, rename it before you install Performance Tools, using the Rename
Object (RNMOBJ) command. This step will ensure the proper operation of the Per-
formance Tools.

Use the following command to place the Performance Tools in library QPFR:
RSTLICPGM LICPGM(5769PT1) DEV(NAME) OPTION(\BASE)

You must then perform one of the following:

Ÿ If you have purchased the Manager feature, use the following command:
RSTLICPGM LICPGM(5769PT1) DEV(tape-device-name) OPTION(1)
Ÿ If you have purchased the Agent feature, use the following command:
RSTLICPGM LICPGM(5769PT1) DEV(NAME) OPTION(2)

If you have several tapes to install, the following situation may occur. After
installing the first tape, you may receive a message saying that the licensed
product is restored but no language objects were restored. If this occurs, load the
next tape and enter the following:
RSTLICPGM LICPGM(5769PT1) DEV(NAME) RSTOBJ(\LNG) OPTION(\BASE)

Another method for installing the Performance Tools product is to type GO LICPGM
and use the menu options.

How Performance Tools Counts Users

Performance Tools is a processor-based product. The usage type is concurrent.
The product is installed with a usage limit *NOMAX.

Printer File and Output Queues

The Performance Tools printer files have a default forms size of 8-1/2 x 11 inches,
an overflow line number of 60, and a characters-per-inch setting of 10 or 15 (this
setting depends on whether the report is 80 or 132 characters wide). If the printer
file characteristics you want are different from the supplied printer file character-
istics, use the Change Printer File (CHGPRTF) command to alter them. Use of the
generic name, QP*, on this command changes all printer files in library QPFR to
the new form size.

 Copyright IBM Corp. 1998 2-1

 The default output queue on the performance job description (QPFRJOBD) is
QPFROUTQ. Reports, submitted as batch jobs, use this job description as the
default. If you want to use a different output queue from the queue established by
Performance Tools, use the Change Job Description (CHGJOBD) command.
Specify the output queue you want to use for the OUTQ parameter on the
CHGJOBD command.

Start Performance Tools (STRPFRT) Command

Use the STRPFRT command to start Performance Tools. After you enter the
command, the IBM Performance Tools menu for your Manager feature or Agent
feature appears. From this display, you can either choose one of the menu
selections, or enter a command:

à ð
PERFORM IBM Performance Tools for AS/4ðð
System: ABSYSTEM
Select one of the following:

1. Select type of status

2. Collect performance data
3. Print performance report
4. Capacity planning/modeling
5. Performance utilities
6. Configure and manage tools
7. Display performance data
8. System activity
9. Performance graphics
1ð. Advisor

7ð. Related commands

Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant

F16=System main menu
á ñ
Press F3 (Exit) or F12 (Cancel) to exit the IBM Performance Tools menu.

Enter commands on the command line. Use F4 (Prompt) and F9 (Retrieve) to

prompt for or retrieve commands that you enter on the command line.

To review any messages that are returned to you on the message line, position the
cursor on the message line and press the Help key for additional detail. Pressing
F10 (Display messages in job log) from this detail display allows you to view all of
the messages currently in the job log.

Each time you use STRPFRT, the following occurs:

Ÿ The library QPFR is added to the library list (between the system and user
positions of the library list).
Ÿ The IBM Performance Tools menu appears.

When you finish using Performance Tools, press F3 (Exit). When you do so, the
library QPFR is removed from the job’s library list.

2-2 Performance Tools V4R2

 Once you use the STRPFRT command to start the Performance Tools, any further
attempt to use the command from within the operating environment for Perform-
ance Tools fails. If you try to start the Performance Tools program when it is
already operating from your job, a message appears that indicates that the oper-
ating environment for Performance Tools is already active. Multiple jobs may use
Performance Tools at the same time but only one performance monitor (data col-
lection job) can be active at any given time.

Displaying the System or Job Status—Manager Feature

If you choose option 1 (Select type of status) on the IBM Performance Tools menu,
the Select Type of Status display appears:

à ð
Select Type of Status to Display

Select one of the following:

1. Work with system status

2. Work with subsystem
3. Work with current job
4. Work with submitted job(s)
5. Work with specified job(s)
6. Work with active jobs
7. Work with disk status

On the Select Type of Status display, you can use a set of OS/400 commands to
provide you with information about the performance of the system or a particular
job.

Each option on the Select Type of Status display has a corresponding command
associated with it, as shown in the following list. To use a function, such as working
with the system status, either enter option 1 on the command line of the Select
Type of Status display or enter WRKSYSSTS on any command line.

Table 2-1. Type of Status Option with Corresponding Command

Type of Status Option Corresponding Command
Work with system status WRKSYSSTS
Work with subsystem WRKSBS
Work with current job WRKJOB
Work with submitted job(s) WRKSBMJOB
Work with specified job(s) WRKJOB
Work with active job(s) WRKACTJOB
Work with disk status WRKDSKSTS

Chapter 2. Starting Performance Tools 2-3

2-4 Performance Tools V4R2
 Chapter 3. Collecting System Performance Data
| The Performance Tools program uses data that is collected by the OS/400
| program. OS/400 gathers data about system performance by using a batch job that
| is called the performance monitor. The performance monitor tracks the activity on
| the system and collects relative data. After the performance monitor collects data,
| the monitor provides a set of files that contain data about the performance of the
| system. In a distributed AS/400 client/server environment, this data can be col-
| lected on managed (or remote, distributed) systems. You can then send the data to
| the central site system where the skills and the tools exist to analyze the collected
| data.

This chapter describes how to collect data using the Start Performance Monitor
(STRPFRMON) command. For the Manager feature, other ways of collecting data
using Performance Tools are described in Chapter 10, Performance
Utilities—Manager Feature, and Chapter 6, System Activity. The figures shown in
the sections following “Summary of Data Collection and Report
Commands—Manager Feature” on page 3-13 show the Performance Tools data
collection commands, and describe when you use each in analyzing the perform-
ance of your system.

The STRPFRMON command is used by both the Agent feature and the Manager
feature. The STRPFRMON command is important in the overall analysis of your
system. Use it to collect data about resources that influence the performance of
your system (processing unit, main storage, auxiliary storage, and communi-
cations).

The STRPFRMON command is provided with the OS/400 program. The Work Man-
agement book contains additional information on collecting performance data and
the database files that result and database file descriptions.

| Why Collect Performance Data

| Collecting data is really the first step toward improving performance. It is a way to
| capture the performance status of the system, or set of systems, involved in getting
| your work done. The collection of data provides a context, or a starting point, for
| any comparisons and analysis that can be done later. When you use your first data
| collections, you have a benchmark for future improvements and a start on
| improving your performance today. Performance problem analysis often begins
| with a simple question: “What changed?” Performance data helps you answer this
| question.

When to Collect Performance Data

You should collect performance data regularly to establish a record of system per-
formance. If you know the performance characteristics of your system, you can
judge the effect of a change in system workload before you make the change.

Use the STRPFRMON command to collect sample data to establish this perform-
ance history. With the Manager feature, use the PRTSYSRPT command to produce
a report from this data. The report provides an overview of the performance vari-

 Copyright IBM Corp. 1998 3-1

 ables that should be tracked in the system’s history. To see an example of the
System Report, see “System Report” on page 7-17.

With the Agent feature, use the Advisor and Work with Historical Data options to
analyze and summarize the performance data collected. See Appendix E, Man-
aging AS/400 System Performance in a Network, for an example.

Collect sample data regularly, so you can make valid comparisons of information.
Collect the data during similar system workloads.
| Note: The performance reports have no exact restrictions on the amount of per-
| formance data that they can process. However, it is recommended that a
| collection be no longer than one week in length. In fact, the optimal col-
| lection strategy is to run a new collection every day.

Also collect sample data before and after major changes, such as installing a new
application. This data provides valuable information on changes to system resource
utilizations as a result of the workload changes. When you review this data, limit
your observations to the differences in the active workloads on the system. Look at
the utilizations for the processing unit, disk, main storage, and communications. If
any of these resources is consistently overcommitted, you should determine the
reason.

To determine the reason for consistent overcommitment of a given resource, collect

trace data. The performance monitor trace data provides more detailed information
on the jobs, programs, and individual transactions that run on the system. If you are
using the Manager feature, depending on the resource, you might use some of the
additional analysis tools. See Chapter 10, Performance Utilities—Manager Feature.

With the Manager feature you can use the PRTTNSRPT command, as described in
“Transaction Report” on page 7-36, to specify the trace data you want to see. You
can determine the elements of transaction response time from these reports. In
addition, you can identify the level of resource use for the measured transactions.

For both the Manager feature and the Agent feature, the length of time you collect
data depends on whether you collect only sample data, or a combination of sample
and trace data. The length of time you collect data can generally be longer when
you collect only sample data. When you collect sample data, you want it to reflect
the changes that occur when the various interactive and batch workloads go from
light to moderate to heavy activity, and back to light activity again.

When you collect trace data, you may want to reduce the length of time for the
measurement, because the area where the trace data is stored is limited in size,
and can contain between 40 000 and 60 000 transactions. The trace collection
ends automatically when this area is filled. Thus, the length of time you collect data
depends on the activity on your system and the volume of trace records that result.
However, because you collect trace data to observe the high utilization of a given
resource, you should try to collect data for workload periods that correspond to
those in which you have seen excessive utilization occur. If you trace data for a
longer period of time, increase the size of the component trace table.

3-2 Performance Tools V4R2

 When to End Data Collection
Use the End Performance Monitor (ENDPFRMON) command to end the data col-
lection. The collected data can automatically be transferred from the collection area
into a database file specified on the DMPTRC parameter on the ENDPFRMON
command. The collected data can be processed by a user-written exit program
specified on the EXITPGM parameter on the ENDPFRMON command. If you are
collecting only sample data, no additional system overhead is associated with stop-
ping the monitor. You can stop the monitor at any time. If you are collecting trace
data, there is additional system overhead and you should try to schedule the end of
the measurement to coincide with a lighter workload period, or use the Dump Trace
(DMPTRC) command to transfer the data at a later time. For example, schedule the
performance monitor collection to end at noon if there is a heavy workload meas-
urement during the morning.

Collecting Sample or Trace Data

| The performance monitor provides for the collection of sample data and trace data.
| Trace Trace data is detailed, and can be collected when detailed applications
| or job analysis is required. Trace data is collected as it happens for
| each transaction and results in a large amount of very detailed data that
| is useful in problem analysis. When you collect trace data, it places
| additional demands on your AS/400 system.
| Use the trace parameter on the STRPFRMON command to control the
| collection of trace data. Trace data resides in the QAPMDMPT database
| file.
| Sample Also called summary data or system data, this data is collected for
| normal trend analysis and performance analysis. The data relates to the
| following:
| Ÿ All jobs on the system
| Ÿ Devices attached to the system
| Ÿ Storage pools
| Ÿ Communications I/O processors
| Ÿ Disk I/O processors
| Ÿ Local workstation I/O processors
| Ÿ Workstation response times
| Sample data is collected at system, resource, job, and device levels and
| on an interval basis. The default collection interval is 15 minutes, but the
| interval can range from 5 - 120 minutes. This means that a performance
| data record is produced for each job and resource on the system at
| each interval. For example, once every 15 minutes.

You control the time between samples by using the interval parameter on the
STRPFRMON command. Sample data is collected and stored in the following data-
base files:

Chapter 3. Collecting System Performance Data 3-3

 Table 3-1 (Page 1 of 2). Database Files that Store Sample Data
File Type of Data
QAPMAPPN Advanced peer-to-peer networking data
QAPMASYN Asynchronous (ASYNC) data
QAPMBSC Binary synchronous communicatios (BSC) data
QAPMBUS Bus counter data
QAPMCIOP Communications controller data
QAPMCONF System configuration data
QAPMDBMON 1 Monitoring data
QAPMDDI Distributed data interface (DDI) data
QAPMDIOP Storage device controller data
QAPMDISK Direct access storage device (DASD) data
| QAPMECL Token-ring LAN protocol data for asynchronous transfer mode (ATM) ports that
| support token-ring LAN emulation and for token-ring ports.
| QAPMETH Ethernet statistics data for asynchronous transfer mode (ATM) ports that
| support Ethernet LAN emulation and for Ethernet ports
QAPMFRLY Frame relay data
QAPMHDLC High-level data link control (HCLC) data
QAPMHDWR Hardware configuration data
QAPMIDLC ISDN data link control (IDLC) B-channel data
QAPMIOPD Communications processor performance data
QAPMJOBS Job data
QAPMLAPD ISDN D-Channel statistics data
QAPMLIOP Local workstation controller (WSC) data
| QAPMMIOP Multifunction controller data. This file includes processing time for token-ring
| and Ethernet LAN emulation.
QAPMPOOL Main storage data
QAPMRESP Local workstation response time data
QAPMRWS Remote workstation response time data
| QAPMSAP Service access point (SAP) data for token-ring, Ethernet, DDI, and frame relay
| ports. The file also reports data for asynchronous transfer mode ports that
| support token-ring and Ethernet LAN emulation.
QAPMSBSD Subsystem description
QAPMSNA SNA performance measurements data
QAPMSNADS SNA distribution services (SNADS) data
QAPMSTND DDI station counter data
| QAPMSTNE Ethernet link station data for asynchronous transfer mode ports that support
| Ethernet LAN emulation and for Ethernet ports
| QAPMSTNL Token-ring LAN link station data for asynchronous transfer mode ports that
| support token-ring LAN emulation and for token-ring ports
QAPMSTNY Frame relay station counter data
QAPMSYS System data

3-4 Performance Tools V4R2

Table 3-1 (Page 2 of 2). Database Files that Store Sample Data
File Type of Data
QAPMTSK Task related performance data
QAPMX25 X.25 data
Note:
1 QAPMDBMON is an optional file that is created or updated when the start database monitor
parameter (STRDBMON) is *YES on the STRPFRMON command.

Whenever you use the STRPFRMON command, you collect sample data, but you
can also collect trace data. You generally choose to collect trace data to gain addi-
tional detailed information about specific jobs and transactions. By collecting trace
data, you can often gain insight into other problems involving job contention,
program resource use, transaction delays, and so on.

The Manager feature allows you to use the Print System Report (PRTSYSRPT)
and Print Component Report (PRTCPTRPT) commands to print the sample data
you collect. To review examples of these reports, see “System Report” on
page 7-17 and “Component Report” on page 7-25.

With the Manager feature, you can use the Print Transaction Report
(PRTTNSRPT), the Print Lock Report (PRTLCKRPT), and the Print Trace Report
(PRTTRCRPT) commands to see the data collected through trace. Refer to “Trans-
action Report” on page 7-36 and to “Lock Report” on page 7-63 to review the infor-
mation provided from trace data collection.

For the Manager feature, some of the commands described in Chapter 10, Per-
formance Utilities—Manager Feature , make use of trace data collected using the
STRPFRMON command. See “Summary of Data Collection and Report
Commands—Manager Feature” on page 3-13 for more information on the com-
mands that use the trace data.

For the Agent feature, you will need to use the Performance Tools Manager feature
to analyze trace data. See Appendix D, Comparison of Performance Tools , for
more information.

For the Manager feature, the Performance Tools program has additional functions
to analyze performance data, including printing of performance reports and perform-
ance utilities. See Appendix D, Comparison of Performance Tools, for more infor-
mation.
Note: You may receive unpredictable results using Performance Tools if the per-
formance monitor is running and a Performance Tools CL command is
using the same member that is collected at that time.

To free up disk space used to save performance data you no longer need, you can
use the Delete Performance Data (DLTPFRDTA) command.

The remaining sections in this chapter describe when and how you collect data
using the STRPFRMON command.

To collect sample or trace data, follow these steps:

Chapter 3. Collecting System Performance Data 3-5

 1. Enter the Start Performance Tools (STRPFRT) command on any command line
to show the IBM Performance Tools menu.

à ð
PERFORM IBM Performance Tools for AS/4ðð
System: ABSYSTEM
Select one of the following:

1. Select type of status

2. Collect performance data
3. Print performance report
4. Capacity planning/modeling
5. Performance utilities
6. Configure and manage tools
7. Display performance data
8. System activity
9. Performance graphics
1ð. Advisor

7ð. Related commands

Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant

F16=System main menu
(C) COPYRIGHT IBM CORP.
á ñ

2. Choose the Collect performance data option on the IBM Performance Tools
menu, and press the Enter key. The Collect Performance Data display appears.

à ð
Collect Performance Data RCHASRðD
ð5/26/95ð7:ð7:27

Performance monitor status:

Status . . . . . . . . . . . . : Not running

Select one of the following:

1. Start collecting data

2. Stop collecting data
3. Work with performance collection

Note: Only one performance monitor function can be active in the system. The
current status of the performance monitor is shown.
3. Choose the Start collecting data option, and press the Enter key. The Start
Collecting Data display appears.

3-6 Performance Tools V4R2

 à ð
Start Collecting Data

Select one of the following:

1. Collect data with defaults

2. Collect data with menus
3. Collect data with command

On this display, there are three ways to start the performance monitor to collect
data. Any option you choose results in the collection of performance data using
STRPFRMON. Options 1 and 2 are designed for a new user of Performance
Tools (they provide more guidance in starting the performance monitor). These
options are discussed in the following sections. If you choose option 3 (Collect
data with command), a prompt for the STRPFRMON command appears as
though you entered the command and pressed F4 (Prompt).

Using System Defaults to Collect Data

If you choose option 1 (Collect data with defaults) on the Start Collecting Data
display, the Collect Data with Defaults display appears.

à ð
Collect Data with Defaults

Type choices, press Enter.

Member . . . . . . . . \GEN Name

Library . . . . . . QPFRDATA Name

Text . . . . . . . . .

Time duration:
Hours . . . . . . . 2 ð-999
Minutes . . . . . . ð ð-6ð

1. Type the name of the member and library where you want to store the perform-
ance data. *GEN creates a member name based on the date and time. The
default for Library is QPFRDATA.
2. Type a description for the sample data in the Text field.
3. Enter the length of time you want to collect performance data in the Time dura-
tion field.
4. Press the Enter key, and the data collection process begins. The defaults for
the other STRPFRMON command parameters are used.
See the CL Reference for more information about the STRPFRMON command
parameters.

Chapter 3. Collecting System Performance Data 3-7

Using Menus to Collect Data
You can change the defaults for the STRPFRMON command by entering the
command directly, or by selecting option 2 (Collect data with menus) on the Start
Collecting Data display.

If you choose option 2, the Collect Data with Menus display appears.

à ð
Collect Data with Menus

Type choices, press Enter.

Member . . . . . . . . \GEN Name

Library . . . . . . QPFRDATA Name

Text . . . . . . . . .

1. Type the member and library name where you want to store the performance
data. The default for Library is QPFRDATA.
2. Enter a description of the performance data in the Text field, if appropriate.
3. Press the Enter key. The Collect Additional Data display appears.

à ð
Collect Additional Data

Type options, press Enter.

1=Collect

Option Type of Data

Trace data
1 Communication data

4. To indicate that you want to collect trace data, type a 1 in the Option column
next to Trace Data, and press the Enter key.
Notes:
a. When this trace starts, all existing traces stop. Only one trace can be active
in the operating system.
b. The Agent feature does not analyze or report on trace data. To analyze or
report on trace data that is collected by the Agent feature, the data must be
sent to a system that has the Manager feature installed.
5. To indicate that you want to collect performance data for communications lines
and objects, type a 1 in the Option column next to Communications Data, and
press the Enter key.
The Set Data Collection Time display appears.

3-8 Performance Tools V4R2

à ð
Set Data Collection Time

Type choices, press Enter.

Number of minutes between

collections . . . . . . 15 5-6ð (increments of 5)

End time option . . . . . 1 1=Elapsed time

2=Time of day
3=No maximum

6. Type how often data should be collected in the Number of minutes between
collections field. Fifteen minutes is usually an adequate sample interval for
problem analysis. This is the sample rate for sample data collection.
7. Type when you want data collection to stop in the End time option field. If you
choose option 1 (Elapsed time), go to step 9. If you choose option 2 (Time of
day) for this prompt, the Set End Time display appears.

à ð
Set End Time

Type choices, press Enter.

Number of days from today . . . . . . . . . . ð ð-9

Time of day . . . . . . . . . . . . . . . . . ðð:ðð:ðð HH:MM:SS

8. Indicate the number of days you want to collect data in the Number of days
from today field. Also indicate when you want data collection to stop in the
Time of day field.
9. If you choose option 1 (Elapsed time) for the End time option prompt on the
Collect Additional Data display, the Set Length of Time to Collect Data display
appears.

à ð
Set Length of Time to Collect Data

Type choices, press Enter.

Length of time to collect data:

Hours . . . . . . . . . 2 ð-999
Minutes . . . . . . . . ð ð-6ð

10. Type the length of time you want to collect data in the Length of time to collect
data fields.

After you press the Enter key, the data collection process begins.

For information on how to print various sample and trace data reports using the
Manager feature, refer to Chapter 7, Performance Reports—Manager Feature.

Chapter 3. Collecting System Performance Data 3-9

Using the STRPFRMON Command to Collect Trace Data
If you choose option 3 (Collect data with command) on the Start Collecting Data
display, the Start Performance Monitor (STRPFRMON) display is shown.

à ð
Start Performance Monitor (STRPFRMON)

Type choices, press Enter.

Member . . . . . . . . . . . . . \GEN Name, \GEN

Library . . . . . . . . . . . . > QPFRDATA Name
Text 'description' . . . . . . . \SAME

Time interval (in minutes) . . . 15 5, 1ð, 15, 2ð, 25, 3ð, 35...
Stops data collection . . . . . \ELAPSED \ELAPSED, \TIME, \NOMAX
Days from current day . . . . . ð ð-9
Hour . . . . . . . . . . . . . . 2 ð-999
Minutes . . . . . . . . . . . . ð ð-99
Data type . . . . . . . . . . . \ALL \ALL, \SYS
Select jobs. . . . . . . . . . . \ALL \ALL, \ACTIVE
Trace type . . . . . . . . . . . \NONE \NONE, \ALL
Dump the trace . . . . . . . . . \YES \YES, \NO
Job trace interval . . . . . . . .5 .5 - 9.9 seconds
Job types . . . . . . . . . . . \DFT \NONE, \DFT, \ASJ, \BCH...
+ for more values
More...
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this disp
F24=More keys

á ñ

Attention

You need to be aware that only one trace can be active in the operating
system. If *ALL is specified on the TRACE parameter while another trace is
running, the other trace is canceled and its data collection is lost.

The parameters that you should consider are:

Ÿ Trace type (TRACE)
Specifies that all of the internal traces that contain performance related informa-
tion are started. If another trace is running at the same time this command
specifying TRACE(*ALL) is issued, the other trace is canceled.
Ÿ Dump the trace (DMPTRC)
Specifies whether or not the trace is dumped when the data collection ends. If
you choose to dump the trace at a later time, you can use the Dump Trace
(DMPTRC) command. Use DMPTRC(*NO) with caution. Do not wait too long to
dump the trace after the monitor has ended. Object addresses in the trace data
resolve to names when DMPTRC is called. Any objects that are deleted
between the time the trace data is collected and the time the trace data is
dumped will not resolve. You will see addresses (which are not useful) instead
of names.

Using DMPTRC(*NO) is illustrated in the following example:

1. On Monday at 8 am, Person A specifies the following:

3-10 Performance Tools V4R2

 STRPFRMON TRACE(\ALL)
MBR(one)
DMPTRC(\NO)
2. On Monday at 2 pm, Person A ends the performance monitor (ENDPFRMON).
3. On Monday at 3 pm, Person B specifies the following:
STRPFRMON TRACE(\ALL)
MBR(two)
Note: When Person B starts to run another trace, the trace data for member
one is destroyed.
4. On Monday at 4 pm, Person B ends the performance monitor (ENDPFRMON).
5. On Monday at 10 pm, Person A dumps the trace data.
DMPTRC MBR(one)

Based on this example, the trace data in MBR(one) will actually be the same as the
trace data in MBR(two). The trace data for MBR(one) will not correspond with the
sample data from MBR(one). For example, PRTSYSRPT(one) and
PRTTNSRPT(one) will contain conflicting information.

Collecting Performance Data Automatically

Note: If you are using automatic data collection for the first time, refer to “Perform-
ance Collection Setup” on page 3-13 before you begin.

You can choose to have your system automatically collect performance data on a
weekly schedule. Automatic performance collection allows you to select specific
days of the week on which automatic data collection is to occur using the OS/400
performance monitor.

On the IBM Performance Tools menu, choose the Collect performance data
option and press the Enter key. The Collect Performance Data display appears.

Chapter 3. Collecting System Performance Data 3-11

 à ð
Collect Performance Data ABSYSTEM
11/ð7/95 ð8:ð7:27
Performance monitor status:
Status . . . . . . . . . . . . : Not running

Select one of the following:

1. Start collecting data

2. Stop collecting data
3. Work with performance collection

Selection or command
===> 3

F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel

á ñ

Choose option 3 (Work with performance collection) and press the Enter key. The
Work with Performance Collection display appears. (You can also use the
WRKPFRCOL command to access the Work with Performance Collection display.)

à ð
Work with Performance Collection

Type options, press Enter.

1=Add 2=Change 3=Hold 4=Remove 5=Display 6=Release

Performance
Opt Collection Status Description
_ __________
_ A RLS
_ Fð9ðð RLS
_ Sð9ðð RLS

Bottom
F3=Exit F5=Refresh F12=Cancel

á ñ

A list of the existing performance collections is displayed. The Performance Col-

lection column specifies the details about when to collect performance data auto-
matically.

3-12 Performance Tools V4R2

 To add a new collection, type a 1 (Add) in the Opt column and press the Enter key.
The Add Performance Collection (ADDPFRCOL) display appears. (You can also
use the ADDPFRCOL command to access the Add Performance Collection
(ADDPFRCOL) display.)

Performance Collection Setup

Automatic performance collection requires a batch job (QPFRCOL) that queries the
schedule created by the Add Performance Collection function and submits the Start
Performance Monitor (STRPFRMON) command at the appropriate times. This job
exists as an autostart job entry in the IBM-supplied subsystems QBASE and QCTL.
The Work Management book contains more information about autostart job entries.

When the job QPFRCOL is started and there are no performance collections
defined that are in a released status, the job ends. Therefore, anytime you add a
collection and there are no existing performance collections in a released status,
start the performance collection as follows:
STRPFRCOL JOBQ(QSYS/QCTL)

You can type this command on any command line. Performance data is collected
until all the performance collections are removed or held, or until you enter
ENDPFRCOL. You need to start performance collection again if you add performance
collections later.

Summary of Data Collection and Report Commands—Manager Feature

Table 3-2 through Table 3-5 in the following sections present the commands for
various levels of data collection. These figures also show the related report com-
mands, show the type of data collected, provide a summary of the information con-
tained in the reports, and describe when you might use these commands.

Refer to the figures indicated for information on the following data collection levels:
Ÿ System ( Table 3-2)
Ÿ Job ( Table 3-3 on page 3-16)
Ÿ File use and structure ( Table 3-4 on page 3-16)
Ÿ Application ( Table 3-5 on page 3-17)

If you use the Performance Tools menus and displays to collect data and produce
reports, these figures may help you understand, at a glance, the capabilities of Per-
formance Tools. If you bypass the menus and displays by entering commands on
the available command entry lines, these figures may serve as a reference for the
available commands.

System-Level Analysis—Manager Feature

System-level data collection and analysis provides you with a comprehensive view
of how the system operates. This information ranges from a system operational
overview to an analysis of individual transactions. System-level data collection and
analysis also provides you with system modeling functions for capacity planning
and performance prediction.

Use system-level data to identify what additional collection and analysis should be
done.

Chapter 3. Collecting System Performance Data 3-13

 A summary of system data collection and report commands is shown in Table 3-2
on page 3-14.

Table 3-2 (Page 1 of 2). System Data Collection (STRPFRMON Command) and Report Commands
Type of Report Information Shown on When to Use the
Level of Data Data Command the Reports Command
Job Disk Sample data ANZPFRDTA Contention analysis and Processing trends
System recommendations System model
Workload
projection
Hardware growth
Processing unit
Main storage
Disk
Job Disk Sample data PRTSYSRPT Workload Utili- Processing trends
System PRTCPTRPT zation System model
Processing unit Workload
Disk projection
Main storage Hardware growth
Communications Processing unit
Model parameters Main storage
External response Disk
times
System Job Trace data PRTTNSRPT Workload Utili- Workload
Program zation projection
Processing unit Hardware growth
Exceptional waits Pool configuration
Transaction detail Overcommitment
Top ten reports Application design
Object contention File contention
Concurrent batch Transaction
jobs Significance
System model Classification
parameters Program use
Transaction summary System model
and detail Processing trends
System Job Trace data PRTTRCRPT Resources used Progression of batch jobs
Program Files Exceptions traced through time
Disk State transitions
Job Program Sample data STRBEST System performance Before installing
Files Disk projections When growth
Capacity planning is anticipated,
Configuration either in
planning hardware or
workload
When a new
application is
to be installed
Performance
analysis

3-14 Performance Tools V4R2

Table 3-2 (Page 2 of 2). System Data Collection (STRPFRMON Command) and Report Commands
Type of Report Information Shown on When to Use the
Level of Data Data Command the Reports Command
Job Program Trace data PRTLCKRPT File, record, or To reduce or
Files Disk object contention remove object
by: contention
Object name Problem analysis
Holding or
requesting job
Time
Job Program Sample data PRTJOBRPT Utilization Problem analysis
Files Disk Processing unit
Disk
Communications
Workload
Job Program Sample data PRTPOLRPT Utilization Problem analysis
Files Disk Main Storage
Workload
Subsystem
Job Program Sample data PRTRSCRPT Utilization Problem analysis
Files Disk I/O Processing
unit
Disk
External response
times
Application or Statis- PRTPEXRPT Program and When general
Program tics procedure performance
Profile statistics on monitoring
Trace calls, CPU usage cannot find
and I/O problems
Sampling of CPU Problem analysis
usage of program
and procedure
instructions
Detailed record
of performance
related events
as they
occurred

For more information on the report commands shown in this figure, see Chapter 7,
Performance Reports—Manager Feature. The Start BEST/1 (STRBEST) command
is described in the BEST/1 Capacity Planning Tool book. The PRTLCKRPT
command is described in “Lock Report” on page 7-63.

Job Trace Analysis—Manager Feature

Job trace analysis enhances the operating system’s standard trace job reports and
provides a summary of job operation and transaction processing. The primary use
for job trace analysis is to determine application flow. You can determine what parts
of a job use the most resources, and measure the effect of program changes rela-
tive to previous trace data. Do not use job trace analysis to determine accurate job
or transaction processing times.

Chapter 3. Collecting System Performance Data 3-15

 A summary of job trace data collection and report commands is shown in Table 3-3
on page 3-16.

For more information about the data collection or report commands, see
Chapter 10, Performance Utilities—Manager Feature.

Table 3-3. Job Trace Data Collection (STRJOBTRC Command) and Report Commands
Information Shown on When to Use the
Level of Data Type of Data Report Command the Reports Command
Job Program Trace data PRTJOBTRC Program name For program devel-
Files ENDJOBTRC Control flow opment
I/O operations To identify jobs or
Full/shared opens programs that
Exceptions perform poorly
Message handling
Disk I/O summary

File Use and Database Structure Analysis—Manager Feature

The commands shown in Table 3-4 provide an overview of the program file use
and the database file structure of an application.

The following contain information for analyzing file use database structure:
Ÿ “Analyze Program (ANZPGM) Command” on page 10-9
Ÿ “Analyze Database File (ANZDBF) Command” on page 10-11
Ÿ “Analyze Database File Keys (ANZDBFKEY) Command” on page 10-13

Table 3-4. File Use and Structure Data Report Commands

Information Shown on When to Use the
Level of Data Report Command the Reports Command
Program File use structure ANZPGM Program file For application use anal-
ysis
Program File use structure ANZDBF Physical file structure For application analysis
Program File use structure ANZDBFKEY Logical file structure For file analysis

Job Analysis
Job analysis provides you with a view of the operational environment for all jobs, or
a group of jobs, in the system at a given time. Use the information from a specific
process analysis to improve the performance of the process. This analysis can help
you improve the program environment to reduce the number of the following:
Ÿ Open files
Ÿ File buffer and work space sizes
Ÿ File open placement in a program
Ÿ Active programs

A summary of job data collection and report commands is shown in Table 3-5 on
page 3-17.

3-16 Performance Tools V4R2

 For more information about the DSPACCGRP command, see “Display Access
Group (DSPACCGRP) Command” on page 10-16.

Table 3-5. Process Data Collection (DSPACCGRP Command) and Report Commands
Information Shown on When to Use the
Level of Data Report Command the Reports Command
Job Program Files DSPACCGRP File use Reduce program size
ANZACCGRP Files used at the Reduce number of
same time open files
Open Data Path Reduce process
Buffer size access group (PAG)
Formats (size I/O
and number) Determine group job
I/O counts candidates
Duplicates
PAG size
Active programs

Chapter 3. Collecting System Performance Data 3-17

3-18 Performance Tools V4R2
Chapter 4. Advisor
The advisor provides an easy-to-use way to improve many of the performance
characteristics of your system.

The advisor fits into the set of Performance Tools between automatic system tuning
and the more specialized tools provided in Performance Tools (such as the Start
Best/1 (STRBEST) command) and the reports (such as a Print System Report).
Appendix D, Comparison of Performance Tools, provides more information about
the functions provided in Performance Tools.

Automatic system tuning is a useful method for maintaining the basic conditions for
good performance. If it is set to work at each IPL, it resets the basic tuning values
to the recommended settings for the system configuration and controlling sub-
system. Dynamic automatic system tuning adjusts only pool sizes and activity levels
of shared pools based on system activity as measured at user-specified intervals.
To adjust the system, the tuner uses a guideline that is calculated based on the
number of jobs.

The advisor can help you to define specific tuning values and other parts of a proc-
essing environment to provide better performance for specific processing conditions
on your system.

The advisor analyzes performance data you collect with the performance monitor
and it can produce recommendations and conclusions to help improve perform-
ance. The advisor might recommend changes to basic system tuning values, and
might list conclusions about conditions that could cause performance problems.

You can choose to have the advisor change system tuning values as it recom-
mends, or you can decide to make only the changes you select. You can use the
advisor’s conclusions to make changes to your system, to guide further perform-
ance data collection, or to help you request performance reports containing more
information and explanations.

The advisor can help you to improve system performance, but it will not identify or
correct all performance problems. The performance information analyzed includes:
Ÿ Storage pool sizes
Ÿ Activity levels
Ÿ Disk and CPU utilization
Ÿ Communications utilizations and error rates
Ÿ Input/output processor utilization
Ÿ Unusual job activities—exceptions or excessive use of system resources
Ÿ Interactive trace data (when available) (Manager feature)

The advisor does not:

Ÿ Make any recommendations for changing specific application programs to
improve their performance
Ÿ Analyze noninteractive trace data

 Copyright IBM Corp. 1998 4-1

 The advisor is a good first tool to use to improve system performance. In many
instances, it will be the only tool required to make the improvements you need. This
chapter takes you through the process for using the advisor. In general, this
process consists of the following steps:
1. Identify when the performance problems occur.
2. Use the performance monitor to collect data during the time periods when each
problem seems to occur.
3. Request the advisor to analyze the data.
4. Use the advisor’s output to change system tuning values, to guide further data
collection, or to request other more detailed performance reports.
5. Observe the effects of any tuning changes, and decide if another cycle through
this process is required to further improve performance or to eliminate
unwanted side effects.

Notes:
1. The examples in this chapter show how to use the advisor, but they do not
contain specific solutions for any performance problems that might exist on a
particular AS/400 system.
2. Sometimes an analysis of data collected during normal system operation can
help in selecting the advisor recommendations to implement to solve perform-
ance problems occurring at other times.
3. At times the advisor will suggest additional analysis using tools available only in
the Manager feature.
4. When the advisor makes no significant recommendations or conclusions and
the system's performance remains unacceptable, analysis at the application
level is required. In this case, the advisor has ruled out many tuning, communi-
cations, and disk problems.
5. When making recommendations, the advisor takes into consideration some
guidelines and threshold values from the BEST/1 hardware table.

Collecting the Right Performance Data

Before collecting performance data, you should clearly describe the problem to be
investigated. From system users’ comments or your own experience, you can begin
to formulate a description of the problem. The problem description does not need to
be overly detailed or technical, just try to simply describe one problem. For
example:
Ÿ Interactive (or batch) processing seems too slow.
Ÿ File updating should go faster.
Ÿ At times the entire system seems to be sluggish.

Next, determine when the problem is most likely to occur. Maybe interactive work is
slow first thing in the morning. Perhaps batch processing seems slow late in the
afternoon.

When you can clearly describe the problem and have determined when it seems to
occur, you are ready to collect performance data to be analyzed by the advisor.

4-2 Performance Tools V4R2

 If possible, focus on collecting data for one problem at a time. Of course, try to
collect the data when the problem seems to occur most often. You can decide later
how much of the data you want the advisor to analyze. For more information about
when to collect performance data and how much to collect, see the first few pages
of Chapter 3, Collecting System Performance Data.

The performance monitor is used to collect performance data. It can be run using
the default data collection values. Information about how to use the performance
monitor is in Chapter 3, Collecting System Performance Data.

Requesting an Analysis
After performance data is collected for the time periods near when a problem
seems to occur, you request the advisor to analyze all or part of that data.

To start the advisor, you can select the Advisor option on the IBM Performance
Tools menu, or type the Analyze Performance Data (ANZPFRDTA) command on
any command line.
Note: To analyze performance data from a library other than QPFRDATA when
using the ANZPFRDTA command, type the command and press F4
(Prompt) to change the library name.

à ð
PERFORM IBM Performance Tools for AS/4ðð
System: ABSYSTEM
Select one of the following:

1. Select type of status

2. Collect performance data
3. Print performance report
4. Capacity planning/modeling
5. Performance utilities
6. Configure and manage tools
7. Display performance data
8. System activity
9. Performance graphics
1ð. Advisor

7ð. Related commands

Selection or command
===> 1ð

F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant

F16=System main menu
(C) COPYRIGHT IBM CORP.
á ñ
The next two steps in requesting a performance data analysis are:
Ÿ Select the member containing the performance data to analyze.
Ÿ Select the time intervals of data to analyze.

Chapter 4. Advisor 4-3

Selecting a Member
When the Advisor option is selected, or the ANZPFRDTA command is run, the
Select Member for Analysis display appears.

à ð
Select Member for Analysis

Library . . . . QPFRDATA

Type option, press Enter.

1=Select 5=Display

Option Member Text Date Time

Q95ð221115 11/ð1/98 22:ð1:ð7
Q983611411 12/27/98 14:11:25
Q98346ð7ð8 12/12/98 ð7:ð8:43
Q9834116ð4 12/ð7/98 16:ð4:23
Q98341142ð 12/ð7/98 14:2ð:22
Q983411411 12/ð7/98 14:11:ð9

BOTTOM
F3=Exit F12=Cancel F15=Sort by name F16=Sort by text
F19=Sort by date/time
(C) COPYRIGHT IBM CORP. 1981, 1998.
á ñ
To request an analysis, select only one member that contains performance data
collected during a time when the problem occurred. When you select a member
and press the Enter key, a Select Time Intervals to Analyze display appears.

Notes:
1. When you return to the Select Member for Analysis display, the 1 typed for the
member remains. This is a reminder that you may want to display this member.
2. When the monitor is running and using one of the members shown in the
Select Member for Analysis display, this member may appear with blank Date
and Time fields until the first interval is collected.

Selecting Time Intervals

4-4 Performance Tools V4R2

 à ð
Select Time Intervals to Analyze

Member . . . . . . . : Q95ð221115 Library . . . . . . : QPFRDATA

Type options, press Enter.

1=Select
Transaction --CPU Util--- --High--- -Pool Fault- Excp
Opt Date Time Count Rsp Tot Int Bch Dsk Unit Mch Usr ID Util
11/ð5 15:29 1ð41 3.1 24 15 6 14 ððð5 2 31 ð4 1
11/ð5 15:59 1ð9ð .2 8 4 2 4 ðð15 ð 6 ð2 ð
11/ð5 16:28 1ð34 .3 7 4 2 5 ððð8 ð 9 ð2 ð
11/ð5 16:58 1412 .3 12 6 4 6 ððð7 ð 11 ð2 ð
11/ð5 17:28 842 .5 1ð 5 3 6 ððð5 ð 8 ð2 ð
11/ð5 17:58 457 .4 6 2 2 3 ððð1 ð 4 ð2 ð
11/ð5 18:28 221 .2 3 1 ð 2 ðð12 ð 1 ð2 ð
11/ð5 18:58 286 .2 4 1 1 2 ððð2 ð 2 ð2 ð
11/ð5 19:28 145 .3 3 1 1 2 ððð7 ð 3 ð2 ð
11/ð5 19:58 1 1.ð ð ð ð 1 ððð1 ð 1 ð2 ð
11/ð5 2ð:28 ð .ð ð ð ð 1 ððð1 ð ð ð2 ð
More...
F3=Exit F5=Refresh F11=Display histogram F12=Cancel F13=Select all
F14=Deselect all

á ñ
The Select Time Intervals to Analyze display lists all the time intervals of perform-
ance data collected in the library member selected on the Select Member for Anal-
ysis display. To analyze a different member, press F12 (Cancel) to return to the
Select Member for Analysis display.

The columns on the Select Time Intervals to Analyze display can help you focus
the analysis on time intervals when the suspected performance problem seems to
have occurred. If there are no obvious reasons to select only some of the displayed
time intervals, you can select them all for analysis by pressing F13 (Select all).

When one or more time intervals are selected for analysis, press the Enter key to
request the analysis by the advisor.
| Note: The Transaction Count field does not include the number of DDM I/Os that
| were generated. Use the Display Performance Data (DSPPFRDTA)
| command to display the value for the logical database I/O for DDM jobs.

Using a Histogram
Sometimes a graph of the data for one of the performance values in the data
makes it easier to select specific time intervals of data for analysis. To define and
display a graph (called a histogram), press F11 (Display histogram) on the Select
Time Intervals to Analyze display. The display then changes to include the Select
Histogram window.

Chapter 4. Advisor 4-5

 à ð
Select Time Intervals to Analyze
............................................
: Select Histogram : ary . . . . . . : QPFRDATA
: :
: Type option, press Enter. :
: 1=Select :
: :
: Opt View : --High--- -Pool Fault- Excp
: Transaction count : Dsk Unit Mch Usr ID Util
: Transaction response time : 22 ðð19 3 13 ð4 6
: Total CPU utilization : 27 ðð26 2 9 ð4 7
: 1 Interactive CPU utilization : 29 ðð19 1 11 ð5 7
: Batch CPU utilization : 21 ðð12 1 13 ð4 7
: High disk utilization : 22 ðð16 2 14 ð4 6
: More... : 17 ðð16 2 9 ð7 5
: F3=Exit F12=Cancel : 18 ððð9 1 14 ð5 3
: : 15 ðð26 1 13 ð4 4
:..........................................: 21 ðð19 2 14 ð7 5
11/ð1 12:ð6 389 1.ð 96 25 63 19 ðð19 1 13 ð7 11
11/ð1 12:11 281 1.3 93 16 68 23 ðð19 3 18 ð7 5
More...
F3=Exit F5=Refresh F11=Display histogram F12=Cancel F13=Select all
F14=Deselect all

á ñ
The View column lists the performance values that can be selected to define the Y
(vertical) histogram axis. The X (horizontal) histogram axis always shows the time
intervals contained in the member.

As an example, to make it easier to see the time intervals where CPU utilization is
the highest, you could select one of the CPU utilization views. A sample histogram
for Interactive CPU Utilization follows:

à ð
Select Time Intervals from Histogram

Type a '1' under each interval to select, press Enter.

Interactive CPU utilization

48 :
44 :
4ð :\ \
36 :\ \
32 :\ \ \ \
28 :\ \ \\ \\ \
24 :\\ \ \ \\\ \\\ \
2ð :\\\\\ \ \ \\\\ \\\ \
16 :\\\\\\ \\\\\\\\\ \\\\\\
12 :\\\\\\\\\\\\\\\\\\\\\\\
8 :\\\\\\\\\\\\\\\\\\\\\\\
4 :\\\\\\\\\\\\\\\\\\\\\\\
ð :\\\\\\\\\\\\\\\\\\\\\\\
11111111111111111111111
+-------+-------+-------+-------+-------+-------+-------+-------+
11:21 12:ð1 12:41
F3=Exit F5=Refresh F11=Display histogram F12=Cancel F13=Select all
F14=Deselect all

á ñ
On this example, it is easy to see and select the time intervals of greatest interac-
tive processing unit use. The number 1 is entered to select each time interval to be

4-6 Performance Tools V4R2

 analyzed. All of the intervals can be quickly selected by pressing F13 (Select all),
as shown in the example.

After the Enter key is pressed on the Select Time Intervals to Analyze display or on
the Select Time Intervals from Histogram display, the advisor analyzes the perform-
ance data for the selected time intervals.

Notes:
1. An analysis of large amounts of performance data can take a long time and
could affect system performance for other users.
2. The analysis performed by the advisor includes all of the types of performance
data for the selected time intervals, and is not limited to the type of data
selected to create the histogram.

Analyzing Trace Data

The advisor can analyze interactive transactions when the performance monitor is
run with the option TRACE *ALL for the selected member. The file QTRTSUM,
produced from the *FILE option of the Transaction report, is analyzed. If the file
does not already exist, the advisor creates QTRTSUM using the default options.
Otherwise, the existing QTRTSUM file is processed.

The command CHGJOBTYP can be run to change the job type of noninteractive
jobs to interactive. After the job types have been changed, the *FILE option of the
Transaction report can be run so that the advisor analyzes the jobs listed as inter-
active.

The performance information analyzed from trace data includes:

Ÿ Exceptions by job
Ÿ Transactions with long seize/lock wait times
Ÿ Unusual transaction activities—excessive wait times

The default is analyzing trace data when it is available. To avoid analyzing trace
data, use the ANZPFRDTA command, press F4 (Prompt), and press F10 (Addi-
tional parameters) to change the value of the DATATYPE parameter to *SAMPLE.
Note: Caution should be used when analyzing an existing QTRTSUM file. The file
may not include time intervals that match the intervals that were picked for
the advisor to analyze.

Using the Advisor’s Results

Depending on the content of the selected performance data, the advisor can
produce recommendations, conclusions, and interval conclusions. What these are
and how you can use them are explained as you look at the following examples.

When a performance data analysis has completed, the Display Recommendations

display shows the results.

Chapter 4. Advisor 4-7

 à ð
Display Recommendations
System: ABSYSTEM

Member . . . . . . . : Q98137ð843 Library . . . . . . : QPFRDATA

System . . . . . . . : ABSYSTEM Version/Release . . : 4/ 2.ð

Start date . . . . . : 11/17/98 Model . . . . . . . : ðð -ð142

Start time . . . . . : ð8:44:ð7 Serial number . . . : 1ð-11AðD

QPFRADJ . . . . . . : ð QDYNPTYSCD . . . . . : 1

Type options, press Enter.

5=Display details

Option Recommendations and conclusions

Recommendations
Decrease pool size for listed pools.
5 Increase pool size for listed pools.
Decrease activity level in listed pools.
Add more main storage.
ASP space capacity exceeded guideline of 8ð.ð%.
Conclusions
Pool fault rates exceeded guideline.
Pool fault rates below guideline.
SDLC utilizations exceeded 5ð% guideline.
More...
F3=Exit F6=Print F9=Tune system F12=Cancel F21=Command line

á ñ

Understanding Recommendations
The Recommendations section of this display deals with conditions that significantly
affect system performance. The recommendations result from comparing the
system values and conditions in the analyzed performance data to the basic
AS/400 performance guidelines.

The recommendations suggest changes to the basic system tuning values that can
improve performance. They also list problems that can be solved by other actions.
In this example, the recommendations about changing pool sizes can be carried out
by changing system tuning values. But, the recommendation about ASP (auxiliary
storage pool) space capacity might require redefining the use of system disk space
or adding to system disk capacity. You might need technical assistance to complete
this type of recommendation. Auxiliary storage pool can be one or more storage
units defined from the disk units or disk unit subsystems that make up auxiliary
storage. ASPs provide a means of isolating certain objects on specific disk units to
prevent the loss of data due to disk media failures on other disk units.

To see more details about a recommendation, type 5 in the Option column. As an

example, the following displays show the details for the example recommendation
Increase pool size for listed pools.

4-8 Performance Tools V4R2

à ð
Display Detailed Recommendation

Recommendation:

Increase pool size for listed pools.

Detailed recommendation:
PFR2567
Technical description . . . . . . . . : The following table shows the
pool identifier, the current pool size, and the suggested pool size.

Pool From To Pool From To

1 1ð238 12193

Increasing the pool size will reduce the page fault rate which will
More...
Press Enter to continue.

F3=Exit F12=Cancel

á ñ
In this example only pool 1 should be increased in size. The text beginning at the
bottom of this display and continuing on the following displays discusses the effects
of changing a pool’s size.

à ð
Display Detailed Recommendation

Recommendation:

Increase pool size for listed pools.

Detailed recommendation:
improve the response time and throughput of jobs in this pool.

Decreasing the pool size will free storage that may in turn be given to
pools with high fault rates.

Removing a pool will free storage that may in turn be given to pools with
high fault rates.

A pool will be increased by at least ten percent of its current size.

Pools that are decreased will all be decreased by the same percentage,
with ten percent of the current size as the maximum amount of decrease.
For example, if a 15ððK pool needs storage, and a 2ðððK and 1ðððK pool can
More...
Press Enter to continue.

F3=Exit F12=Cancel

á ñ
Many recommendations include this type of information to help you choose the right
changes to make to your system.

Chapter 4. Advisor 4-9

Changing System Tuning Values
To see and select the tuning changes related to the recommendations, press F9
(Tune system) on the Display Recommendations display. A display similar to the
following appears:

à ð
Select Tuning Recommendations

Advisor Current Data

Name/ Recommended System Collection
Value To Be Changed Number Value Value Value
POOLSIZE (K) \MACHINE 12193 942ð 11ð85
ACTIVITY LEVEL \BASE 6 7 6
POOLSIZE (K) \INTERACT 7ð755 39683 7ð755
ACTIVITY LEVEL \INTERACT 27 21 31
POOLSIZE (K) \SPOOL 8ð 49 8ð
ACTIVITY LEVEL \SPOOL 3 2 3

Bottom
Select one of the following:

1. Tune to advisor’s recommendations

2. Restore system to data collection values
Selection

F3=Exit F12=Cancel

á ñ
On this Select Tuning Recommendations display you have several choices:
Ÿ Select menu option 1 (Tune to advisor’s recommendations) to have the advisor
make all the changes shown in the Advisor Recommended Value column.
Usually this is a good choice to make when starting to solve a performance
problem.
Ÿ Leave the values as they are listed in the Current System Value column.
Ÿ Select menu option 2 (Restore system to data collection values) to have the
advisor set the values as they were when the analyzed performance data was
collected (shown in the Data Collection Value column).
Ÿ Write down the tuning values that fit your needs, and use the appropriate
system commands to change the values individually.

Notes:
1. The analysis and recommendations are based on the Data Collection Values.
The Current System Value column is there for your reference and in case you
want to reset your configuration to what it was at the time of data collection. If
the Advisor Recommended Value equals the Data Collection Value, then the
advisor is saying that this is an adequate setting for the workload analyzed. If
the Advisor Recommended Value does not equal the Data Collection Value,
then you will see recommendations and conclusions as to what should be
changed.
2. When the dynamic tuning support is active (the system value is 2 or 3), the
storage pool sizes and activity levels are automatically changed. Because of
this automatic change, the advisor is unable to process the tuning request.

4-10 Performance Tools V4R2

Understanding Conclusions
The Conclusions section of the Display Recommendations display lists conditions
that could have affected performance when the analyzed data was collected. These
conditions can include thresholds reached, save and restore activities, transmission
line errors, and so on.

à ð
Display Recommendations
System: ABSYSTEM
Member . . . . . . . : Q98137ð843 Library . . . . . . : QPFRDATA
System . . . . . . . : ABSYSTEM Version/Release . . : 4/ 2.ð
Start date . . . . . : ð5/17/98 Model . . . . . . . : ðð -ð142
Start time . . . . . : ð8:44:ð7 Serial number . . . : 1ð-11AðD
QPFRADJ . . . . . . : ð QDYNPTYSCD . . . . . : 1

Type options, press Enter.

5=Display details

Option Recommendations and conclusions

Recommendations
Decrease pool size for listed pools.
Increase pool size for listed pools.
Decrease activity level in listed pools.
Add more main storage.
ASP space capacity exceeded guideline of 8ð.ð%.
Conclusions
5 Pool fault rates exceeded guideline.
Pool fault rates below guideline.
SDLC utilizations exceeded 5ð% guideline.
More...
F3=Exit F6=Print F9=Tune system F12=Cancel F21=Command line

á ñ
Some conclusions describe conditions that caused the advisor to make particular
recommendations. Other conclusions not related to recommendations can be used
as guides for collecting more performance data, or for adjusting the system.

To see more details about a conclusion, type 5 in the Option column. The following
example is the display showing details for the conclusion Pool fault rates exceeded
guideline that supports the recommendation to increase the size of pool 1.

Chapter 4. Advisor 4-11

 à ð
Display Detailed Conclusion

Conclusion:

Pool fault rates exceeded guideline.

Detailed conclusion:
PFR2513
Technical description . . . . . . . . : The following table shows the
pool identifier, the maximum fault rate over all the intervals, the fault
rate guideline, the number of intervals the guideline was exceeded out of
1 intervals, and the date and time the maximum fault rate occurred. For
pool 2 (\BASE) the guideline is based on the fact that there are no user
jobs running in \BASE.

ID Rate Guide Intervals Date Time

1 3.6 3.ð 3 11/ð1/98 12:31:ð4

More...
Press Enter to continue.

F3=Exit F12=Cancel

á ñ
In this example, the guideline of 3 faults was exceeded for pool 1 in three of the
analyzed time intervals. The maximum fault rate was 3.6.

Understanding Interval Conclusions

The Interval Conclusions section of the Display Recommendations display contains
the detailed data to support the conclusions for the analyzed time intervals.

à ð
Display Recommendations
System: ABSYSTEM
Member . . . . . . . : Q98137ð843 Library . . . . . . : QPFRDATA
System . . . . . . . : ABSYSTEM Version/Release . . : 4/ 2.ð
Start date . . . . . : ð5/17/98 Model . . . . . . . : ðð -ð142
Start time . . . . . : ð8:44:ð7 Serial number . . . : 1ð-11AðD
QPFRADJ . . . . . . : ð QDYNPTYSCD . . . . . : 1

Type options, press Enter.

5=Display details

Option Recommendations and conclusions

Interval Conclusions
5 Pool fault rates above guideline.
Total disk I/O was 225. (179 Reads and 46 Writes)
No performance problems found on listed TRLAN lines.
Total system fault rate above 5ð.ð guideline.
Thrashing may have occurred.

Bottom
F3=Exit F6=Print F9=Tune system F12=Cancel F21=Command line

á ñ
To see more details about an interval conclusion, type 5 in the Option column. The
following example is the display showing details for the sample interval conclusion,

4-12 Performance Tools V4R2

Pool fault rates above guideline, which supports the conclusion that pool fault rates
exceeded the guideline.

à ð
Display Detailed Interval Conclusion

Interval conclusion:

Pool fault rates above guideline.

Detailed interval conclusion:

PFR2553
Technical description . . . . . . . . : The following table shows the
pool identifier, the fault rate, and the time the fault rate exceeded the
guideline.

Id Rate Guide Date Time

1 3.ð 3.ð 11/ð1/98 11:21:13

1 3.1 3.ð 11/ð1/98 12:11:ð6
1 3.6 3.ð 11/ð1/98 12:31:ð4

Bottom
Press Enter to continue.

F3=Exit F12=Cancel

á ñ
In this example we see exactly when, and by how much, the fault rate guideline
was exceeded for pool 1 in the analyzed time intervals.

Following is the details display for another type of sample interval conclusion, Total
disk I/O was 225:

à ð
Display Detailed Interval Conclusion

Interval conclusion:

Total disk I/O was 225. (179 Reads and 46 Writes)

Detailed interval conclusion:

PFR2847
Cause . . . . . : This is the sum of all the disk I/O for all the
selected intervals for all the disk devices. This does not indicate a
problem; this data is provided simply for reference.

Bottom
Press Enter to continue.

F3=Exit F12=Cancel

á ñ

Chapter 4. Advisor 4-13

 An interval conclusion like this one provides information but does not support a
conclusion or recommendation. It does not report a problem but provides informa-
tion that can be helpful in understanding how your system is performing.

Tune System by Advisor’s Recommendations

After you request a performance analysis, and look over the results, often the next
step is to have the advisor tune the system as it recommends. Do this by selecting
menu option 1 (Tune to advisor’s recommendations) on the Select Tuning Recom-
mendations display.

Next, observe the effects of the changes. Use the performance monitor to collect
more performance data during the next time period when you expect the problem to
occur. Also, observe the system and watch for the usual symptoms of the problem.
Ask users who experienced the problem if they still notice it. Watch for any pos-
sible unwanted side effects from the tuning changes. These can occur if the
changes are not fully compatible with some of your processing requirements, or if
several problems are being worked on.

The first attempt to solve a basic performance problem can be successful. But
sometimes the steps described in this chapter must be repeated until the best pos-
sible performance is achieved for your system and your processing requirements.

The original problem may continue or new problems may occur. The advisor might
have no further recommendations or conclusions that you can use. At this time you
could use other performance reports and commands to work on the problem. These
are described in Work Management.

Sometimes tuning alone will not solve performance problems. To handle the
intended work load, a system might need additional main storage, disk storage, or
processing speed. BEST/1 can be used to determine if system processing capaci-
ties should be increased. For more information about BEST/1 and capacity plan-
ning, see the BEST/1 Capacity Planning Tool book.

4-14 Performance Tools V4R2

Chapter 5. Displaying Performance Data
This chapter describes how to interactively view the performance data collected by
the Start Performance Monitor (STRPFRMON) command. This Display Perform-
ance Data function can be used either while the performance monitor is active or
after the collection is completed.
Note: The data collection does not need to contain the trace data in order to use
this display function. Trace data may be required, however, to further
analyze performance problems isolated by this function.

Display Performance Data

To interactively display sample performance data, you can do one of the following:
Ÿ Type the Display Performance Data (DSPPFRDTA) command on any
command line using the default value of *SELECT for the member parameter.
Ÿ Type the DSPPFRDTA command on any command line specifying a member
for the MBR parameter.
Note: If you specify a member on the DSPPFRDTA command, you do not see
the Select Performance Member display or the Select Time Intervals to
Display display. The Display Performance Data function starts to read
the performance database files.
Ÿ If you are using the Manager feature, select the Display performance data
option on the IBM Performance Tools menu.
Ÿ If you are using the Agent feature, select the Advisor option on the IBM Per-
formance Tools menu. Then select option 5 from the next menu.

à ð
Select Performance Member

Library . . . . QPFRDATA

Type option, press Enter.

1=Select

Option Member Text Date Time

_ Q983221324 11/17/98 13:24:ð6
_ Q9831ð1458 11/ð5/98 14:58:2ð
_ Q983ð81ðð9 11/ð3/98 1ð:ð9:13
_ Q983ð7ð759 11/ð2/98 ð7:59:25

Bottom
F3=Exit F12=Cancel F15=Sort by name F16=Sort by text
F19=Sort by date/time
(C) COPYRIGHT IBM CORP. 1981, 1998.
á ñ
The member name, a text description, and the date and time you collected each
set of performance data appear on this display. If you cannot find the data you

 Copyright IBM Corp. 1998 5-1

 want to display, use the Roll keys to page through the list of members or use the
appropriate function key to sort the sets of performance data. You can sort the data
by member name, text description, or by the date and time the member was
created. When you find the performance data you want to display, for the Manager
feature type a 1, or for the Agent feature type a 5, in the corresponding Option field.
Note: When the monitor is running and using one of the members shown in the
Select Performance Member display, this member may appear with blank
Date and Time fields until the first interval is collected.

If you are searching for a member located in a library that is different from the one
currently listed in the Library field at the top of the display, type a new library name
in the field and press the Enter key. A list of the performance members available in
the library you specified appears. You can then select to display one of them.

After you select a performance member to display, the Select Time Intervals to
Display display appears.

à ð
Select Time Intervals to Display

Member . . . . . . . : Q983221324 Library . . . . . . : QPFRDATA

Type options, press Enter.

1=Select

Transaction --CPU Util--- --High--- -Pool Fault- Excp

Opt Date Time Count Rsp Tot Int Bch Dsk Unit Mch Usr ID Util
11/17 13:39 427 2.2 1ð 4 4 4 ððð1 ð 6 ð2 1
11/17 13:54 441 .9 12 7 3 6 ððð5 ð 6 ð2 ð
11/17 14:ð9 16ð .6 6 3 2 4 ððð5 ð 6 ð2 ð
11/17 14:24 189 .5 5 2 1 4 ððð5 ð 6 ð2 ð
11/17 14:39 328 .5 8 3 3 6 ððð5 1 8 ð2 ð
11/17 14:54 167 .5 5 1 3 4 ððð5 ð 5 ð2 ð
11/17 15:ð9 282 .6 8 3 3 4 ðð1ð ð 5 ð2 ð
11/17 15:19 167 .3 7 3 2 5 ððð5 ð 6 ð2 ð

Bottom
F3=Exit F5=Refresh F11=Display histogram F12=Cancel F13=Select all
F14=Deselect all

á ñ
Select the time interval for which you want to display performance data.

The Display Performance Data function then starts to read the performance data-
base files. All the performance information required by this function is processed
now, so there is reasonable response time when moving between displays later.
Note: The initial processing may cause a noticeable delay in presenting the first
display.

After all the data is processed, the main display for the Display Performance Data
function appears.

5-2 Performance Tools V4R2

à ð
Display Performance Data

Member . . . . . . . . Q983221324 F4 for list

Library . . . . . . QPFRDATA

Elapsed time . . . . : ðð:54:16 Version . . . . . . : 4

System . . . . . . . : MACHINE1 Release . . . . . . : 2.ð
Start date . . . . . : 11/17/98 Model . . . . . . . : 2ðð-2ð51
Start time . . . . . : 13:24:ð6 Serial number . . . : 1ð-1288312
QPFRADJ . . . . . . : ð QDYNPTYSCD . . . . . : 1

CPU utilization (priority) . . . . . . . . . . . : 42.ð1

CPU utilization (other) . . . . . . . . . . . . : 34.92
Job count . . . . . . . . . . . . . . . . . . . : 539
Transaction count . . . . . . . . . . . . . . . : 1ð836
Transactions per hour . . . . . . . . . . . . . : 1198ð
Average response (seconds) . . . . . . . . . . . : 1.74
Disk utilization (percent) . . . . . . . . . . . : 5.24
Disk I/O per second . . . . . . . . . . . . . . : 138.9
Logical DB I/Os for DDM jobs . . . . . . . . . . : 52.ð

F3=Exit F4=Prompt F5=Refresh F6=Display all jobs F1ð=Command entry

F12=Cancel F24=More keys

á ñ
On this display you can change both the Member and Library fields. If you type a
new member name in the Member field and press the Enter key, the data in that
member appears on the display. If you type a new library name in the Library field
and press the Enter key, the program tries to locate the member in the specified
library. If you press F4 (Prompt) after you enter the library name, the Select Per-
formance Member display uses the specified library to present a list of data col-
lections.

The Display Performance Data function helps you analyze the performance data. It
highlights the values on this display that exceed the threshold values.

Therefore, if the interactive CPU utilization or the disk utilization exceeds the
threshold, the field is highlighted on the display.

To access a command line after you start the Display Performance Data function,
press F10 (Command entry). This allows you to work from a command entry
display without exiting the display function. Once you exit the command entry, you
are immediately returned to the Display Performance Data display without having to
experience the initial processing delay.

To better understand system performance, you might want to view the data sorted
by category. The second set of function keys on this display allows you to group
the performance data by subsystem, job type, or interval.

Ÿ
Ÿ
Ÿ

F13=Display by subsystem F14=Display by job type F15=Display by interval

F24=More keys

á ñ

Chapter 5. Displaying Performance Data 5-3

 By categorizing the data, you might be able to isolate a group of jobs that require
further analysis. If you do, you can then display the performance statistics for indi-
vidual jobs.

The next sections describe the displays that show the performance data separated
into the subsystem, job type, and interval categories.

Display Performance Data by Subsystem

If you press F13 on the Display Performance Data display, the Display by Sub-
system display appears.

à ð
Display by Subsystem

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter. Press F6 to display all jobs.

5=Display jobs

CPU Job Tns Average Disk

Option Subsystem Util Count Count Response I/O
_ \MACHINE 9.86 13ð ð .ðð 42933
_ QBASE .ðð 1 ð .ðð 29
_ QBATCH 17.99 32 ð .ðð 23337
_ QCTL 2.35 3 295 .71 1ð937
_ QINTER 4ð.28 355 1ð221 1.77 79822
_ QSNADS .ðð ð ð .ðð ð
_ QSPL 5.33 145 ð .ðð 22639

Bottom
F3=Exit F6=Display all jobs F12=Cancel F14=Display by job type
F15=Display by interval

á ñ
This display categorizes the performance data according to the subsystem in which
the activity occurred.

From this display you may be able to isolate a single subsystem or group of sub-
systems that are of particular interest. To view the performance data for the jobs in
particular subsystems, type a 5 in the appropriate Option fields and press the Enter
key. If you do not want to select a particular subsystem, but would rather view the
data for all the jobs in the measurement, press F6 (Display all jobs).

Display Performance Data by Job Type

If you press F14 on the Display Performance Data display, the Display by Job Type
display appears.

5-4 Performance Tools V4R2

 à ð
Display by Job Type

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter. Press F6 to display all jobs.

5=Display jobs

CPU Job Tns Average Disk

Option Job Type Util Count Count Response I/O
_ Autostart .ðð ð ð .ðð ð
_ Batch 19.65 69 ð .ðð 33769
_ Evoke .ðð ð ð .ðð ð
_ Interactive 4ð.94 354 1ð5ð2 1.74 81815
_ LIC 9.66 127 ð .ðð 42347
_ Pass-Through .ð8 3 14 2.42 195
_ Sbs Monitor .97 4 ð .ðð 2984
_ System .2ð 3 ð .ðð 586
_ Writer 4.32 1ð7 ð .ðð 18ðð1

Bottom
F3=Exit F6=Display all jobs F12=Cancel F13=Display by subsystem
F15=Display by interval

á ñ
This display categorizes the performance data according to the job types of the
jobs running on your system.

From this display you may be able to isolate a single job type or group of job types
that are of particular interest. To view the performance data for the jobs of partic-
ular job types, type a 5 in the appropriate Option fields and press the Enter key. If
you do not want to select a particular job type, but would rather view the data for all
the jobs in the measurement, press F6 (Display all jobs).

Display Performance Data by Interval

If you press F15 on the Display Performance Data display, the Display by Interval
display appears.

Chapter 5. Displaying Performance Data 5-5

 à ð
Display by Interval

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter. Press F6 to display all jobs.

5=Display jobs

CPU Job Tns Average Disk

Option Date Time Util Count Count Response I/O
_ ð9/19/98 14:ð1:48 71.35 212 957 1.7ð 12644
_ ð9/19/98 14:ð6:47 82.68 22ð 982 1.87 12893
_ ð9/19/98 14:11:47 87.27 229 955 1.96 14717
_ ð9/19/98 14:16:46 63.ð5 236 921 1.83 11693
_ ð9/19/98 14:21:46 55.34 2ð7 823 2.11 9715
_ ð9/19/98 14:26:45 54.79 2ð4 885 2.54 9585
_ ð9/19/98 14:31:45 49.17 2ð9 8ð6 1.37 9787
_ ð9/19/98 14:36:44 58.54 215 1ð21 1.31 11ð73
_ ð9/19/98 14:41:45 71.72 241 12ð7 1.56 151ð3
_ ð9/19/98 14:46:44 71.ð4 222 1ð89 1.6ð 15ð18
More...
F3=Exit F6=Display all jobs F12=Cancel F13=Display by subsystem
F14=Display by job type

á ñ
This display categorizes the performance data according to the collection intervals
that occurred during the measurement.

From this display, you may be able to isolate a single interval or group of intervals
that are of particular interest. To view the performance data for the jobs in particular
intervals, type a 5 in the appropriate Option fields and press the Enter key. If you
do not want to select a particular interval, but would rather view the data for all the
jobs in the measurement, press F6 (Display all jobs).

Display Jobs
If you selected a subsystem on the Display by Subsystem display, selected a job
type on the Display by Job Type display, selected an interval on the Display by
Interval display, or pressed F6 (Display all jobs) on any of these or the Display
Performance Data display, the Display Jobs display appears.

5-6 Performance Tools V4R2

 à ð
Display Jobs

Subsystem . . . . : QINTER Member . . . . . : Q983221324

Elapsed time . . : ðð:54:16 Library . . . . : QPFRDATA

Type options, press Enter.

5=Display job detail

Job CPU Tns Avg Disk

Option Job User Number Type Util Count Rsp I/O
_ DSPð1 QSECOFR ð2222ð INT 1.15 2ð2 1.2 941
_ DSP1ðð QPGMR ð22213 INT .7ð 155 1.4 694
_ DSP89 QPGMR ð22219 INT .64 75 3.6 674
_ DSPð2 QSYSOPR ð22222 INT .63 127 2.2 956
_ DSP47 QPGMR ð2221ð INT .62 164 1.1 888
_ DSPð9 QSECOFR ð22215 INT .59 129 1.2 661
_ DSP57 QPGMR ð22223 INT .56 151 2.ð 625
_ DSP38 QPGMR ð22217 INT .56 52 2.7 2177
_ DSP92 QPGMR ð222ð7 INT .55 163 2.3 7ð6
_ DSPð3 QSYSOPR ð22168 INT .52 96 1.8 1235
More...
F3=Exit F12=Cancel F15=Sort by job F16=Sort by job type
F19=Sort by CPU F24=More keys

á ñ
This display appears when you request to view the jobs in a particular subsystem.
If you request a job type or interval, the Subsystem indicator at the top of the
display is replaced by a Job Type or a Interval indicator. Also, if you selected a
particular job type, the Job Type column does not appear because all the jobs have
the same type as indicated by the Job Type field at the top of the display. If you
request to see all the jobs (by pressing F6 on the Display by Subsystem, the
Display by Job Type, or the Display by Interval displays) the appropriate indicator
(Subsystem, Job Type, or Interval) appears at the top of the display showing a
value of ‘*ALL’ and the Job Type column is present. If F6 is pressed from the
Display Performance Data display, there is no indicator, such as subsystem, job
type, or interval, at the top of the display. Also, in this case, the Job Type column
would be present.

Display Job Detail

If you type a 5 in the Option field next to a job on either the Display Jobs display or
the Display Remote Jobs display (see page 5-12), and press the Enter key, the
Display Job Detail display appears.

Chapter 5. Displaying Performance Data 5-7

 à ð
Display Job Detail

Job . . . . . . . . : DSPð1 Job type . . . . . . : INT

User . . . . . . . : QSECOFR Subsystem . . . . . : QINTER
Number . . . . . . : ð2222ð Pool . . . . . . . . : ð4
Member . . . . . . . : Q983221324 Priority . . . . . . : 2ð
Library . . . . . : QPFRDATA Elapsed time . . . . : ðð:54:16

CPU Tns Average Disk Act-> Wait-> Act->

Interval Seconds Count Response I/O Wait Inel Inel
14:ð1:48 .ððð ð .ð ð .ð .ð .ð
14:ð6:47 .ððð ð .ð ð .ð .ð .ð
14:11:47 .883 2 3.ð 293 .8 .ð .ð
14:16:46 4.93ð 18 2.8 138 3.6 .ð .ð
14:21:46 12.935 43 1.7 116 8.6 .ð .ð
14:26:45 1.938 25 .4 67 5.ð .ð .ð
14:31:45 .693 1ð .5 28 2.ð .ð .ð
More...
Press Enter to continue.

F3=Exit F11=View 2 F12=Cancel F15=Sort by interval

F24=More keys

á ñ
The Display Job Detail display provides you with the performance data for a partic-
ular job, broken down by collection intervals. This display presents the performance
information using three different views, which can be accessed by function keys.
F11 shows you the next view in the series.

Display Performance Data for System Resources

When you are on the Display Performance Data display, you may want to view the
performance data specifically related to storage pools, disk units, or communi-
cations lines, instead of the job-related information previously discussed. The third
set of function keys, as shown below, allows you to do this.

Ÿ
Ÿ
Ÿ

F19=Display pool detail F2ð=Display disk detail

F21=Display communications detail F24=More keys

á ñ

Display Pool Detail

If you press F19 on the Display Performance Data display, the Display Pool Detail
display appears.

5-8 Performance Tools V4R2

 à ð
Display Pool Detail

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter.

5=Display pool intervals

Size Act Tns Avg Expert

Opt Pool (K) Lvl Count Rsp Cache
_ ð1 22528 ð ð .ð ð
_ ð2 4ð96 32767 449 3.2 ð
_ ð3 3ð72 5ð ð .ð ð
_ ð4 6ð416 72 1ð387 1.6 ð
_ ð5 8192 3ð ð .ð ð

Bottom
F3=Exit F11=Display faults and pages F12=Cancel F15=Sort by pool
F24=More keys

á ñ
The Display Pool Detail display presents performance information for each pool in
the measurement. Two views are used in order to present all the pool information.

Although the Display Pool Detail display presents the pool information as totals for
the entire measurement, you may want to examine the data for a particular pool
over time. Using the Display pool intervals option allows you to view the same pool
information broken down into the time intervals in which it occurred.

Display Pool Interval

By typing a 5 in the Option field next to a pool and pressing the Enter key, the
Display Pool Interval display appears with performance information for that pool.

Chapter 5. Displaying Performance Data 5-9

 à ð
Display Pool Interval

Pool . . . . . . . . : ð4 Member . . . . . . . : Q983221324

Elapsed time . . . . : ðð:54:16 Library . . . . . : QPFRDATA

Size Act Tns Avg DB DB Non-DB Non-DB

Interval (K) Lvl Count Rsp Faults Pages Faults Pages
14:ð1:48 6ð416 72 895 1.7 2.7 135.ð 12.7 56.8
14:ð6:47 6ð416 72 915 1.9 2.2 176.3 13.ð 52.7
14:11:47 6ð416 72 936 1.9 2.4 119.9 14.7 6ð.7
14:16:46 6ð416 72 886 1.8 2.1 167.7 14.8 76.6
14:21:46 6ð416 72 811 2.1 2.6 237.6 13.2 54.5
14:26:45 6ð416 72 861 1.4 2.ð 1ð5.3 13.8 93.6
14:31:45 6ð416 72 794 1.3 2.1 125.2 12.7 53.6
14:36:44 6ð416 72 942 1.3 2.1 154.ð 13.6 54.ð
14:41:45 6ð416 72 1146 1.5 2.7 159.5 16.4 69.4
14:46:44 6ð416 72 1ð77 1.6 2.5 157.5 16.ð 64.7
14:51:44 6ð416 72 1124 1.4 2.8 136.2 15.6 2ð9.2
Bottom
Press Enter to continue.

F3=Exit F11=Display transitions F12=Cancel F15=Sort by interval

F24=More keys

á ñ
The Display Pool Interval display presents the same columns of information as the
Display Pool Detail display, except that the data is broken down by time intervals. A
second view (not shown here) also exists for the Display Pool Interval display,
which presents the data for the state transitions.

Display Disk Detail

If you press F20 on the Display Performance Data display, the Display Disk Detail
display appears.

à ð
Display Disk Detail

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter.

5=Display disk intervals

----Activity Per Second-----

Unit Size ASP % % Read Read Write Write
Opt Unit Name (M) ID Used Busy Rqs (K) Rqs (K)
_ ððð1 DDðð1 427 ð1 6ð.7 6.5 1.ð 2.3 1.8 1.6
_ ððð2 DDðð2 427 ð1 6ð.8 5.5 .8 2.2 1.7 1.1
_ ððð3 DDðð3 427 ð2 68.5 .5 .ð .ð .2 .3
_ ððð4 DDðð4 427 ð2 68.5 .4 .ð .ð .2 .2
_ ððð5 DDðð5 427 ð1 6ð.8 6.8 1.2 4.6 1.8 1.3
_ ððð6 DDðð6 427 ð1 6ð.8 6.9 1.3 4.4 1.7 1.2
_ ððð7 DDðð7 427 ð1 6ð.8 6.8 1.2 4.ð 1.9 1.4
_ ððð8 DDðð8 427 ð2 68.5 .5 .ð .ð .2 .3
_ ððð9 DDðð9 427 ð1 6ð.8 7.4 1.2 4.1 2.ð 1.5
More...
F3=Exit F11=View 2 F12=Cancel F15=Sort by unit F22=Sort by % used
F23=Sort by % busy

á ñ

5-10 Performance Tools V4R2

 The Display Disk Detail display presents performance information for each disk unit
attached to the system on which the data collection was performed.

Although the Display Disk Detail display presents the disk information as totals for
the entire measurement, you may want to examine the data for a particular disk
unit over time. Using the Display disk intervals option allows you to view the same
disk information broken down into the time intervals in which it occurred.

Display Disk Interval

By typing a 5 in the Option field next to a disk unit and pressing the Enter key, the
Display Disk Interval display appears with performance information for that disk
unit.

à ð
Display Disk Interval

Unit . . . . . . . . : ððð1 Member . . . . . . . : Q983221324

Size (M) . . . . . . : 427 Library . . . . . : QPFRDATA
Unit name . . . . . : DDðð1 Elapsed time . . . : ðð:54:16

-------Activity Per Second-------

ASP % % Read Read Write Write
Interval ID Used Busy Rqs (K) Rqs (K)
14:ð1:48 ð1 6ð.6 7.8 1.4 2.6 2.1 1.8
14:ð6:47 ð1 6ð.6 5.7 .8 1.4 1.5 1.3
14:11:47 ð1 6ð.6 4.7 .7 1.8 1.9 1.3
14:16:46 ð1 6ð.6 4.5 .9 2.7 1.2 .9
14:21:46 ð1 6ð.6 5.3 .7 1.6 1.4 1.1
14:26:45 ð1 6ð.7 4.ð .8 1.7 .7 .7
14:31:45 ð1 6ð.7 5.1 .9 2.ð 1.4 1.3
14:36:44 ð1 6ð.7 5.3 .7 2.ð 1.6 1.8
14:41:44 ð1 6ð.7 12.8 1.5 4.2 3.3 3.1
More...
Press Enter to continue.

F3=Exit F12=Cancel F15=Sort by interval F22=Sort by % used

F23=Sort by % busy

á ñ
The Display Disk Interval display presents the same columns of information as the
Display Disk Detail display, except that the data is broken down by time intervals.
Note: The Size (M) field is at the top of the display because the size of the disk
unit cannot change from one interval to the next.

Display Communications Line Detail

If you press F21 on the Display Performance Data display, the Display Communi-
cations Line Detail display appears.
Note: The data in the following figures are examples only and should not be used
to tune any system.

Chapter 5. Displaying Performance Data 5-11

 à ð
Display Communications Line Detail

Member . . . . . . . : Q983221324 Elapsed time . . . . : ðð:54:16

Library . . . . . : QPFRDATA

Type options, press Enter.

5=Display remote jobs 7=Display communications interval data

Line Line Line Tns Average Job %

Option ID Type Speed Count Response Count Busy
_ BSC41 BSC 2.4 ð .ðð ð .ð
_ BSC2ð BSC 2.4 ð .ðð ð .ð
_ SDLCLINEA1 SDLC 9.6 13ð1 1.8ð 56 52.ð
_ SDLC1ðð SDLC 9.6 1ð47 1.22 71 3ð.6
_ SDLCB SDLC 9.6 558 1.16 28 2ð.3
_ SDLCLINEð2 SDLC 9.6 1168 2.85 45 44.8
_ SDLCLINEð3 SDLC 9.6 1ð24 1.34 34 3ð.3
_ SDLC15ðð SDLC 9.6 1556 1.74 45 58.7
_ SDLC1ððð SDLC 9.6 1317 2.35 96 5ð.4

Bottom
F3=Exit F12=Cancel F15=Sort by line ID F2ð=Sort by transactions
F24=More keys

á ñ
Figure 5-1. Display Communications Line Detail

The Display Communications Line Detail display presents performance information

for each communications line attached to the system.

The Display Communications Line Detail display presents the totals for each line in
the measurement. One of the options on this display lets you view performance
data for the jobs using a communications line. The other option displays the time
interval performance data for a communications line.

Display Remote Jobs

If you type a 5 (Display remote jobs) on the Display Communications Line Detail
display in the Option field next to a communications line and press the Enter key,
the Display Remote Jobs display appears with the performance information for that
line listed by job.

5-12 Performance Tools V4R2

 à ð
Display Remote Jobs

Line . . . . . . . . : SDLCLINEA1 Member . . . . . . . : Q983221324

Line type . . . . . : SDLC Library . . . . . : QPFRDATA
Line speed . . . . . : 9.6 Elapsed time . . . . : ðð:54:16

Type options, press Enter.

5=Display job detail

Job CPU Tns Avg Disk

Option Job User Number Type Util Count Rsp I/O
_ DSP92 QPGMR ð222ð7 INT .55 163 2.3 7ð6
_ DSP79 QPGMR ð22191 INT .52 71 2.6 734
_ DSP13 QPGMR ð22ð75 INT .5ð 1ð4 2.2 776
_ DSP41 QSECOFR ð22123 INT .36 1ð8 1.6 7ð4
_ DSP63 QPGMR ð222ð3 INT .31 46 2.2 785
_ DSP85 QSECOFR ð22193 INT .26 58 2.3 527
_ DSPð8 QSYSOPR ð2212ð INT .21 53 2.1 397
_ DSP54 QPGMR ð22175 INT .2ð 51 1.7 275
_ DSP5ð QPGMR ð22172 INT .17 41 1.9 261
More...
F3=Exit F12=Cancel F15=Sort by job F16=Sort by job type
F19=Sort by CPU F24=More keys

á ñ
Figure 5-2. Display Remote Jobs

If you type a 5 in the Option column, you can display more detailed information for
the remote job. This option calls the Display Job Detail display, just as option 5 did
from the Display Jobs display. Refer to “Display Job Detail” on page 5-7 for infor-
mation on the performance data that will be shown.

Display Communications Interval Data

To see a display of performance data for a communications line by time interval,
type a 7 (Display communications interval data) in the Option field next to the com-
munications line on the Display Communications Line Detail display, and press the
Enter key. The resulting Display Communications Interval Data display lists the per-
formance averages and totals for that communications line for the time intervals in
the current performance data member.

From the Display Communications Interval Data display you can request data about
the jobs using the communications line during any of the listed time intervals. To do
this, type a 5 in the Option column by the selected time interval.

Each communications protocol has its own type of Display Communications Interval
Data display, but all are quite similar. An example and description for synchronous
data link control (SDLC) is shown in Figure 5-3 on page 5-14. Other communi-
cations protocols are:
Ÿ X.25
Ÿ Token-ring LAN area network (TRLAN)
Ÿ Ethernet local area network (ELAN)
Ÿ Distributed data interface (DDI)
Ÿ Frame relay (FRLY)
Ÿ Binary synchronous communications (BSC)

Chapter 5. Displaying Performance Data 5-13

 Ÿ Asynchronous data link control (ASYNC)
Note: Option 7 (Display communications interval data) is not valid for IDLC lines.
To view ISDN and IDLC line information, press F13 (Display network inter-
face data).

à ð
Display Communications Interval Data

Line ID . . . . . . . . : MPLSCHI Member . . . . . . . . . : MONDAY

Line type . . . . . . . : SDLC Library . . . . . . . : QPFRDATA
Line speed . . . . . . . : 19.2 Elapsed time . . . . . . : ðð:24:5ð
IOP name . . . . . . . . : CMBð1

Type options, press Enter,

5=Display remote jobs

Pct Pct
I Frames Frames
Itv Line I Frames Trnsmitd Frames Recd
Option End Util Trnsmitd in Error Recd in Error
13:ð8:ðð 78 1,818 ð9 1,818 ð2
13:23:ðð 78 1,818 ð7 1,818 ðð
13:38:ðð 78 1,818 ðð 1,818 ðð
13:53:ðð 78 1,818 ðð 1,818 ðð
14:ð8:ðð 78 1,818 ðð 1,818 ðð
Bottom
F3=Exit F11=View 2 F12=Cancel F15=Sort by itv end
F2ð=Sort by line util F24=More keys

á ñ
Figure 5-3. Display Communications Interval Data for SDLC

Display Remote Interval Jobs

This display lists information about the jobs using a communications line during a
time interval. To request it, type a 5 (Display remote jobs) in the Option column by
a time interval on a Display Communications Interval Data display, and press the
Enter key. The following display appears:

5-14 Performance Tools V4R2

 à ð
Display Interval Remote Jobs

Interval . . . . . . . . : 13:ð8 Member . . . . . . . . . : MONDAY

Line ID . . . . . . . . : MPLSCHI Library . . . . . . . : QPFRDATA
Line type . . . . . . . : SDLC Elapsed time . . . . . . : ðð:24:5ð
Line speed . . . . . . . : 19.2
Line utilization . . . . : 78%

Job CPU Tns Avg Disk

Job User Number Type Util Count Rsp I/O
DSP15 Xð7733 ð3ð191 DDM .16 19 .ð 23ð
DSP4ð SMITH ð3ð275 INT .24 24ð 3.5 1598
DSP43 U5531 ð3ð212 DDM .ðð ð .ð 76

Bottom
F3=Exit F12=Cancel F15=Sort by job F16=Sort by job type
F19=Sort by CPU F24=More keys

á ñ
The end time for the selected time interval, the line name, line type, line speed, and
average use during the time interval are shown in the fields Interval, Line ID, Line
type, Line speed and Line utilization at the top of this display. The column
descriptions are the same as for Figure 5-2 on page 5-13.

Display Network Interface Data

To see a display of performance data for Integrated Services Digital Network
(ISDN) network interfaces, press F13 (Display network interface data) on the
Display Communications Line Detail display.
Note: F13 is shown only if your data collection contains ISDN data.

This display shows performance information for each ISDN network interface and
channel pair configured on the system that data was collected for. From this
display, you can view the data on a per-interval basis by typing a 7 by the network
interface and channel you want to see.

More information regarding ISDN can be found in the ISDN Support book.

Chapter 5. Displaying Performance Data 5-15

 à ð
Display Network Interface Data

Member . . . . . . . : ISDN1 Elapsed time . . . . : ð1:54:39

Library . . . . . : V4R2CT

Type options, press Enter.

7=Display channel interval data

Transmit/ Percent Percent

Receive/ Total Frames Total Frames
Network Average Frames Trnsmitd Frames Received
Option Interface Channel Line Util Trnsmitd Again Received in Error
ISDNSS_A B1 ð1/ð1/ð1 8778 ð 88ð2 ð
ISDNSS_B B1 13/17/15 85ð6 7 9859 9
ISDNSS_A B2 ðð/ðð/ðð 3758 ð 3779 ð
ISDNSS_B B2 ðð/ðð/ðð 3779 ð 3736 ð
ISDNSS_A D 11/15/13 1318 4ð 143ð 38
ISDNSS_B D ðð/ðð/ðð 994 ð 993 ð

Bottom
F3=Exit F12=Cancel F15=Sort by network interface F16=Sort by channel
F2ð=Sort by line util F24=More keys

á ñ
Figure 5-4. Display Network Interface Data

Display Channel Interval Data

To see a display of performance data for a specific ISDN network interface and
channel by time interval, type a 7 (Display channel interval data) in the Option field
next to the network interface on the Display Network Interface Data display and
press the Enter key. The resulting Display Channel Interval Data display lists the
performance averages and totals for that network interface and channel for the time
intervals in the current performance data member.

Each channel type has its own type of Display Channel Interval Data display. An
example and description of this display for each channel type follow.

5-16 Performance Tools V4R2

à ð
Display Channel Interval Data

Network Interface . : ISDNSS_B Member . . . . . . . : ISDN1

Channel . . . . . . : B1 Library . . . . . : QPFRDATA
Line speed . . . . . : 64.ð Elapsed time . . . . : ð1:54:39
IOP name . . . . . . : CMBð1

Type options, press Enter.

5=Display remote jobs

Transmit/ Percent Percent

Receive/ Total Frames Total Frames
Itv Line Average Frames Trnsmitd Frames Received
Opt End ID Line Util Trnsmitd Again Received in Error
14:33:11 ISDNSS_B1 99/99/99 347 33 13ð 39
14:36:22 ISDNSS_B1 99/99/99 35 1ðð 75 1ðð
14:44:54 ISDNSS_B1 27/99/7ð 376 52 578 34
14:5ð:55 ISDNSS_B2 ðð/ð1/ðð 256 ð 255 ð
14:56:19 ISDNSS_B2 24/39/31 238 15 286 25
More...
F3=Exit F11=View2 F12=Cancel F15=Sort by itv end F19=Sort by line ID
F2ð=Sort by line util F24=More keys

á ñ
Figure 5-5. Display Channel Interval Data for B-channel

From this display you can request data about the jobs using the communications
line listed during any of the time intervals. To do this, type a 5 in the Option column
by the selected time interval.

à ð
Display Channel Interval Data

Network Interface . : ISDNSS_A Member . . . . . . . : ISDN1

Channel . . . . . . : D Library . . . . . : QPFRDATA
Line speed . . . . . : 16.5 Elapsed time . . . . : ð1:54:39
IOP name . . . . . . : CMBð1

Transmit/ Loss
Receive/ -Outgoing Calls- -Incoming Calls- of
Itv Average Total Percent Total Percent Frame
End Line Util Calls Rejected Calls Rejected Alignment
14:46:2ð 12/21/16 42 28 15 26 452
15:ð1:19 2ð/ð6/13 74 74 33 1ðð 135
15:16:17 ðð/ðð/ðð ð ð 5 ð ð
15:21:17 ðð/ðð/ðð ð ð 2 ð ð
15:31:16 ðð/ðð/ðð ð ð 2 ð ð
15:46:14 ð7/1ð/ð9 21 1ðð 34 1ðð 348
More...
Press Enter to continue.

F3=Exit F11=View 2 F12=Cancel F13=Display maintenance channel

F24=More keys

á ñ
Figure 5-6. Display Channel Interval Data for D-channel

Chapter 5. Displaying Performance Data 5-17

Display Maintenance Channel Data
This display shows performance data for the ISDN maintenance channel. To
request it, press F13 (Display maintenance channel) on the Display Channel
Interval Data for D-channels display.
Note: F13 is shown only if the system you collected data for had its ISDN mainte-
nance channel active.

à ð
Display Maintenance Channel Data

Network Interface . : ISDNSS_A Member . . . . . . . : ISDN1

Line speed . . . . . : 16.5 Library . . . . . : QPFRDATA
IOP name . . . . . . : CMBð1 Elapsed time . . . . : ð1:54:39

Percent
Percent Severely Far
Itv Errored Errored DTSE DTSE End Code
End Seconds Seconds In Out Violation
14:46:2ð 5ð 36 734 83 32
15:ð1:19 6 24 32 14 52
15:16:17 ð ð ð ð ð
15:21:17 ð ð ð ð ð
15:31:16 ð ð ð ð ð
15:46:14 99 99 36 45 66
16:ð1:13 95 8ð 11 9 1
More...
Press Enter to continue.

F3=Exit F12=Cancel F15=Sort by itv end F2ð=Sort by DTSE in

F21=Sort by DTSE out F22=Sort by percent severely errored seconds

á ñ
Figure 5-7. Display Maintenance Channel Data

5-18 Performance Tools V4R2

Chapter 6. System Activity
This chapter describes the functions that allow you to work with performance data
for the jobs and Licensed Internal Code tasks currently running on the system.
These functions provide the ability to interactively view and collect the data in a
QAITMON database file using the Work with System Activity (WRKSYSACT)
command and to print reports based on the collected data (print activity report).
These functions are available as AS/400 commands or through option 8 (System
activity) on the IBM Performance Tools menu. If you select option 8, the System
Activity menu appears.

à ð
System Activity

Select one of the following:

1. Work with system activity

2. Print activity report
Ÿ
Ÿ
Ÿ

Refer to “Work with System Activity” and “Print Activity Report” on page 6-9 for a
description of both selections shown on the System Activity menu.

Work with System Activity

The Work with System Activity (WRKSYSACT) command allows you to view and
collect performance data in a realtime fashion. This data, which consists of CPU
utilizations, synchronous and asynchronous I/O counts, storage amounts, and
more, is reported for any job or task that is currently active on the system.
Note: To be considered active, a job or task must use at least one-tenth of 1%
(.1%) of the processing unit or perform one I/O operation.

Similar to the performance monitor started by the Start Performance Monitor

(STRPFRMON) command, only one call of the Work with System Activity function
can be active at one time. If this function is currently active when the WRKSYSACT
command is run, you receive a message indicating that:
Ÿ The function is already active
Ÿ The name of the user profile who is running the command.

The performance statistics reported by this function represent activity that has
occurred during the elapsed time since a previous collection. Notice that this may
contrast with other system functions that generally provide cumulative values until
specifically reset. In most cases the time interval between data collections ranges
from 1 second to several minutes, depending on how often you want to view or
collect new data. On systems with very little activity, a subsecond refresh interval
may be possible.

 Copyright IBM Corp. 1998 6-1

 Depending on the value specified for the OUTPUT parameter of the Work with
System Activity command, the data gathered by this function is processed in one of
the following methods:
Ÿ Shown on the display station only
Ÿ Written to a database file only
Ÿ Shown on the display station and written to a database file

When the data is written to a database file only, this function submits a batch job
under the name WRKSYSACT. When the data is written to a file and shown on the
display, the statistics are put in the file each time the display is refreshed. This
does not include the data presented on the initial display of the Work with System
Activity display. See the file descriptions in Table 6-1 on page 6-8.

When the data is being directed to the display station (either the first or third
method), the first display that appears resembles the following:

à ð
Work with System Activity
ð2/ð8/96 1ð:45:19

Automatic refresh in seconds . . . . . . . . . . . . . . . . . 5

Elapsed time . . . . : ðð:ðð:ð2 Overall CPU . . . : 74.1

Type options, press Enter.

1=Monitor job 5=Work with job
Total Total
Job or CPU Sync AsyncPAG
Opt Task User Number Thread Pty Util I/O I/OFault
_ DSP25 QPGMR ðð8124 2ð 56.1 12 4 ð
_ DSPð1 QSECOFR ðð8122 2ð 6.4 5 2 ð
_ DSPð7 QSYSOPR ðð8123 2ð 4.4 1 ð ð
_ DSPð4 QSECOFR ðð8119 1 3.7 ð ð ð
_ #AðððA ð 1.ð ð ð ð
_ #5ððð9 ð .6 ð ð ð
_ SMAIð3 ð .3 ð ð ð
_ ROUTð4 ð .3 ð ð ð
_ SMAIð1 ð .3 ð ð ð
_ #Aððð3 ð .3 ð ð ð
More...
F3=Exit F1ð=Update list F11=View 2 F12=Cancel F19=Automatic refresh
F24=More keys

á ñ
Figure 6-1. System with Single Processor

The input-capable field Automatic refresh in seconds at the top of the display con-
trols the amount of time between display refreshes when the automatic refresh
feature is active. Refer to “Automatic Refresh Mode” on page 6-4 for more informa-
tion on this field. The second field at the top of the display, Elapsed time, reflects
the length of time in which the currently shown performance statistics occurred.
Described in a different way, this value represents the time between the last display
refresh and the next-to-last display refresh.
Note: The Work with System Activity display automatically gathers the data twice
before displaying the first display. Therefore, the initial Elapsed time should
be approximately 2 seconds, which means that the statistics shown
occurred in the 2 seconds previous to the current display.

6-2 Performance Tools V4R2

Single-Processor System
The Overall CPU represents the CPU utilization for the entire system during the
elapsed time. This value does not always equal the sum of the individual CPU utili-
zations shown in the list, since a job or task could use an extremely small amount
of processing unit time, thus affecting the overall utilization, but not use enough
CPU resource to be included in the list of active jobs. (Refer to the requirements for
being considered active at the beginning of this chapter.) The discrepancy in CPU
utilizations, however, is small and should have little effect on the usability of this
function.

Also, the Overall CPU could exceed 100% on extremely busy systems, because
the data collection process does not occur instantaneously. However, you should
be aware that overall CPU utilizations slightly over 100% are an acceptable possi-
bility.

Multiple-Processor System
For a multiple-processor system, Overall CPU field is replaced by these fields:
Ÿ Minimum CPU util
Ÿ Maximum CPU util
Ÿ Average CPU util
Ÿ Number of CPUs

For each of the CPU utilization fields, the value shown is the total CPU utilization
divided by the number of processors shown in the Number of CPUs field.
Figure 6-2 shows the Work with System Activity display for a system with more
than one processor:

à ð
Work with System Activity ABSYSTEM
ð2/ð8/96 1ð:45:19
Automatic refresh in seconds . . . . . . . . . . . . . . . . . 5
Elapsed time . . . . . . : ðð:ð1:ð2 Average CPU util . . : 97.ð
Minimun CPU util . . . . : 94.1 Maximum CPU util . . : 99.9
Number of CPUs . . . . . : 4

Type options, press Enter.

1=Monitor job 5=Work with job
Total Total
Job or CPU Sync Async PAG
Opt Task User Number Thread Pty Util I/O I/O Fault
QPADEVððð7 SUSTAITA ð29844 1 25.2 ð ð ð
QPADEVðð12 PATO ð29845 1 24.ð ð ð ð
QPADEVðð15 SOFIACN ð29846 1 25.5 ð ð ð
QPADEVðð6ð BRLEON ð29849 1 24.8 4 212 ð

Bottom
F3=Exit F1ð=Update list F11=View 2 F12=Cancel F19=Automatic refresh
F24=More keys
á ñ
Figure 6-2. System with Multiple Processors

The two options shown on the Work with System Activity allow you to analyze spe-
cific jobs and tasks that appear in the list. Refer to “Monitoring Specific Jobs” on
page 6-5 and “Working with Jobs” on page 6-6 for more information on these
options.

Chapter 6. System Activity 6-3

| The jobs and tasks are presented on this display in decreasing order of a number
| of different methods. This order is initially controlled by the Sequence (SEQ)
| parameter on the Work with System Activity command. The default is to sort the
| jobs and tasks by CPU utilization. Once the function has been started, however,
| F16 (Sequence by xxx) serves as a switch between the sorting methods.

| The Work with System Activity function uses different views to present all the per-
| formance statistics. Pressing F11 shows you the next view in the series and
| pressing F10 refreshes the current view.
Note: As mentioned above, the Job or Task column is shown only when
INFTYPE(*ALL) is specified. This value for the Information type parameter
instructs the function to display both jobs and tasks. Specifying
INFTYPE(*JOBS) causes the Job or Task column, to be replaced by the
Job column because only jobs are to be displayed. Similarly, specifying
INFTYPE(*TASKS) causes the Job or Task column to be replaced by the
Task column since only tasks are to be displayed. Later sections of this
chapter describe how to switch between these information types through the
use of function keys.

Automatic Refresh Mode

Automatic Refresh mode represents an important feature of the Work with System
Activity function. Once started, this mode continually updates the display without
requiring further user intervention.

To start the Automatic Refresh mode, first enter the desired number of seconds
between refreshes in the Automatic refresh in seconds field. This value, which has
an initial default of 5 seconds, can range from a minimum of 1 second to a
maximum of 900 seconds (15 minutes).
Note: Setting the Automatic refresh seconds at 5 or greater generally results in
the Work with System Activity function using reasonably small amounts of
the processing unit, depending on the size of the system being monitored.
Setting this value lower than 5 seconds causes this function to use larger
amounts of the processing unit, and therefore, is not recommended.

Once you have established the desired refresh interval, pressing F19 (Automatic
refresh) starts the automatic refresh mode. Automatic refresh continues to display
the same view and type of information that was previously selected. For example, if
you had been examining both jobs and tasks using View 1, the display appears as
follows.

6-4 Performance Tools V4R2

 à ð
Work with System Activity ABSYSTEM
1ð/ð1/98 ð9:51:48
Automatic refresh in seconds . . . . . . . . . . . . . . . . . 5
Elapsed time . . . . : ðð:ðð:ð2 Overall CPU util . . : 1.6

Type options, press Enter.

1=Monitor job 5=Work with job
Total Total
Job or CPU Sync Async PAG
Opt Task User Number Thread Pty Util I/O I/O Fault
1 QPADEVðððI AABB ð54145 ðððððððC 1 .9 2 ð ð
1 CFINTð1 ð .1 ð ð ð
1 QUMBVROUTR AABBCCDD ð53919 ðððððð19 5ð .1 ð ð ð
1 QCQEPMON QSVMSS ð53243 ððððððð1 5ð .1 ð ð ð
1 QPFRADJ QSYS ð53184 ððððððð1 ð .1 ð ð ð

Bottom
F3=Exit F1ð=Update list F11=View 2 F12=Cancel F19=Automatic refresh
F24=More keys
á ñ
Figure 6-3. System with Single Processor

| The Automatic Refresh function attempts to maintain even refresh intervals by com-
| pensating for the time required to process, display, and, possibly, write the perform-
| ance data. Therefore, you may occasionally notice that the elapsed time does not
| exactly match the value specified for the Automatic refresh in seconds field. Press
| F19 to end the automatic refresh function.

Monitoring Specific Jobs

While using the Work with System Activity function, you may want to view the per-
formance statistics for a set of jobs and tasks on the system. By typing a 1 in the
Opt column before a list entry, that job or task is selected for monitoring. You may
monitor as many as 20 jobs and tasks at a single time. If five jobs and tasks are
selected for monitoring, the Work with System Activity display appears as follows:

Chapter 6. System Activity 6-5

 à ð
Work with System Activity ABSYSTEM
1ð/ð1/978 ð9:51:48
Automatic refresh in seconds . . . . . . . . . . . . . . . . . 5
Elapsed time . . . . : ðð:ðð:ð2 Overall CPU util . . : 1.6

Type options, press Enter.

1=Monitor job 5=Work with job
Total Total
Job or CPU Sync Async PAG
Opt Task User Number Thread Pty Util I/O I/O Fault
1 QPADEVðððI AABB ð54145 ðððððððC 1 .9 2 ð ð
1 CFINTð1 ð .1 ð ð ð
1 QUMBVROUTR AABBCCDD ð53919 ðððððð19 5ð .1 ð ð ð
1 QCQEPMON QSVMSS ð53243 ððððððð1 5ð .1 ð ð ð
1 QPFRADJ QSYS ð53184 ððððððð1 ð .1 ð ð ð

Bottom
F3=Exit F1ð=Update list F11=View 2 F12=Cancel F19=Automatic refresh
F24=More keys
á ñ
Figure 6-4. System with Single Processor

Once you have selected jobs and tasks for monitoring, the Work with System
Activity function is placed in a subset mode. While in this mode, you see perform-
ance data for only the selected jobs and tasks whenever the display is refreshed.
Also in this mode, you can use option 5 (Work with job) on a job and the job
remains in the selected group. To remove a single job or task from the selected
group (as long as it is not the last or only selected entry), blank out the option field
and press the Enter key. This causes a new group to be built from those entries
that still have a 1 in the Opt field.

To return to normal operating mode, press either F13 (Jobs and tasks), F14 (Jobs
only), or F15 (Tasks only). These function keys are the only way to end the moni-
toring feature without exiting the Work with System Activity function.

Working with Jobs

By typing a 5 in the option field next to a job and pressing the Enter key, the Work
with Job (WRKJOB) command is started for that job. Selecting more than one job
before pressing the Enter key causes the WRKJOB command to be started multiple
times.
Note: Option 5 (Work with job) is valid only with jobs. This function cannot be
started for tasks.

Refer to the CL Reference or the Work Management book for further information on
the Work with Job command.

6-6 Performance Tools V4R2

Displaying Different Information Types
As previously mentioned, you can control the type of information being shown on
the display. This control comes through the use of the INFTYPE (Information type)
parameter or through the use of F13 (Display jobs and tasks), F14 (Display jobs
only), or F15 (Display tasks only).

If you specify INFTYPE(*ALL) on the Work with System Activity command or press
F13, statistics for both jobs and tasks are shown. Column headings and function
keys similar to the following appear on the Work with System Activity display:

Total Total
Job or CPU Sync Async PAG
Opt Task User Number Thread Pty Util I/O I/O Fault
_ DSP25 QPGMR ðð8124 2ð 56.1 12 4 ð
_ DSPð1 QSECOFR ðð8122 2ð 6.4 5 2 ð
_ SMAIð3 ð .3 ð ð ð
_ ROUTð4 ð .3 ð ð ð
.
.
.
F14=Display jobs only F15=Display tasks only F16=Sequence by I/O
F24=More keys

á ñ
If you specify INFTYPE(*JOBS) or press F14, statistics for jobs only are shown.
Column headings and function keys similar to the following appear on the Work
with System Activity display:

Total Total
CPU Sync Async PAG
Opt Job User Number Thread Pty Util I/O I/O Fault
_ DSP25 QPGMR ðð8124 2ð 56.1 12 4 ð
_ DSPð1 QSECOFR ðð8122 2ð 6.4 5 2 ð
.
.
.
F13=Display jobs and tasks F15=Display tasks only F16=Sequence by I/O
F24=More keys

á ñ
And finally, if you specify INFTYPE(*TASKS) or press F15, statistics for tasks only
are shown. Column headings and function keys similar to the following appear on
the Work with System Activity display:

Total Total
CPU Sync Async PAG
Opt Task Thread Pty Util I/O I/O Fault
_ SMAIð3 ð .3 ð ð ð
_ ROUTð4 ð .3 ð ð ð
.
.
.
F13=Display jobs and tasks F14=Display jobs only F16=Sequence by I/O
F24=More keys

á ñ

Chapter 6. System Activity 6-7

Accessing Work Management Functions
To assist you in analyzing the performance of the system, function keys F20
through F23 have been set up to provide access to several Work Management
functions. The third set of function keys appears on the Work with System Activity
display as follows:

F2ð=Work with active jobs F21=Work with system status

F22=Work with subsystems F23=Work with disk status F24=More keys

á ñ
F20 starts the Work with Active Jobs (WRKACTJOB) command. F21 starts the
Work with System Status (WRKSYSSTS) command. F22 starts the Work with Sub-
systems (WRKSBS) command, and F23 starts the Work with Disk Status
(WRKDSKSTS) command. Refer to the Work Management book for further infor-
mation on these commands.

Content of Database File QAITMON

The collected performance data is stored in the file QAITMON located in the library
specified by the LIB parameter on the Work with System Activity command. Each
performance collection, which is stored in a member determined by the MBR
parameter, contains one record for each active job or task in an interval.

Table 6-1 describes the content of a single record in QAITMON.

Table 6-1 (Page 1 of 2). File QAITMON

Field Name Attributes Description
LVLID CHAR(7) The level of the module that collected this data and the level of
this file in the form VVRRRFF, where VV = version number, RRR
= release number, and FF = file level.
DTETIM CHAR(13) The date (CMMDDYY) and time (HHMMSS) that the data was col-
lected.
ITVTIM PACKED(11,0) The time between data collections, where one unit equals 4096
microseconds.
CPUTOT PACKED(11,0) The total processing unit time used by all tasks and jobs during
the interval, where one unit equals 4096 microseconds. For
multiple-processor systems, this is the average use by all
processors.
NAME CHAR(10) The job or task name for this entry.
JOBUSR CHAR(10) The user profile associated with a job.
JOBNBR CHAR(6) The number assigned to the job.
PTY CHAR(3) The priority of the job or task when the data was collected.
CPUDLT PACKED(11,0) The processing unit time used by this task or job during the
interval, where one unit equals 4096 microseconds. For multiple-
processor systems, this is the average use by all processors.
IOTOT PACKED(11,0) The total physical I/O operations (synchronous and asynchronous)
performed by this job or task.
SDBR PACKED(11,0) The number of synchronous database reads.
SNDBR PACKED(11,0) The number of synchronous nondatabase reads.
SDBW PACKED(11,0) The number of synchronous database writes.
SNDBW PACKED(11,0) The number of synchronous nondatabase writes.
ADBR PACKED(11,0) The number of asynchronous database reads.

6-8 Performance Tools V4R2

 Table 6-1 (Page 2 of 2). File QAITMON
Field Name Attributes Description
ANDBR PACKED(11,0) The number of asynchronous nondatabase reads.
ADBW PACKED(11,0) The number of asynchronous database writes.
ANDBW PACKED(11,0) The number of asynchronous nondatabase writes.
PAGFLT PACKED(11,0) The number of process access group faults.
RSRV2 PACKED(11,0) Reserved.
JTFLAG CHAR(1) A flag indicating whether this record represents a job or task
where ‘00’X = job and ‘80’X = task.
RSRV1 CHAR(4) Reserved.
PERMW PACKED(11,0) The number of writes that were for permanent objects.
IOPND PACKED(11,0) The number of I/O-pending page faults.
SMSYNC PACKED(11,0) The number of waits for asynchronous I/O operations to complete.
OVRTOT PACKED(11,0) The total number of binary, decimal, and floating point overflow
exceptions.
CPU11 PACKED(11,0) For multiple-processor systems, the time used in processor one by
jobs and tasks during the interval. One unit of time equals 4096
microseconds.
CPU21 PACKED(11,0) For multiple-processor systems, the time used in processor two by
jobs and tasks during the interval. One unit of time equals 4096
microseconds.
CPUCNT PACKED(3,0) The number of active processors in the system during data col-
lection.
CPU31 PACKED(11,0) For multiple-processor systems, the time used in processor three
by jobs and tasks during the interval. One unit of time equals 4096
microseconds.
CPU41 PACKED(11,0) For multiple-processor systems, the time used in processor four by
jobs and tasks during the interval. One unit of time equals 4096
microseconds.
CPU5–CPU32 PACKED(11,0) For multiple-processor systems, the time used in processor n by
jobs and tasks during the interval. One unit of time equals 4096
microseconds.
| THDID CHAR(8) The thread identifier assigned to a job. When a task is running,
| this field is blank.
| STGALC PACKED(11,0) Reserved.
| STGDLC PACKED(11,0) Reserved.
Notes:
1 Even though these fields are no longer shown on the Work with System Activity display, you
can query the QAITMON file after running WRKSYSACT OUTPUT(\FILE) or WRKSYSACT
OUTPUT(\BOTH).

Print Activity Report

The Print Activity Report (PRTACTRPT) command creates a report using the per-
formance data collected by the Work with System Activity (WRKSYSACT)
command. This report is produced in the spooled file QPITACTR.

Depending on the value specified for the Report Type (RPTTYPE) option on the
Print Activity Report command, one of two report types, or both, are created. The
summary report provides the top 10 listings showing the most CPU-intensive and
the most I/O-intensive entries over the entire specified period. The detailed report
shows a selected number of entries for each interval in the specified period. These
entries are ordered according to a user selected field. Refer to the following
sections for more detail on each of these report types.

Chapter 6. System Activity 6-9

Summary Activity Report
The Summary Activity Report consists of two sections. The first lists (in decreasing
order) the top 10 entries according to CPU utilization during the specified period,
and the second lists (also in decreasing order) the top 10 entries according to total
I/O activity performed during the specified period. The value used for total I/O is
actually the sum of the total synchronous I/O and the total asynchronous I/O. If 10
active jobs or tasks are not present in the specified period, these sections list as
many entries as are available.

The following represents a sample Summary Activity Report:

System Activity Report 3/3ð/98 11:ð6:26
Page 1
Member . . . . : QAITMON Report Type . . . . : SUMMARY Version . . . . : 4 Started . . . . : ð3/3ð/98 1ð:56:22
Library . . . . : QPFRDATA Release . . . . : 2.ð Stopped . . . . : ð3/3ð/98 1ð:57:ð9
Order by CPU Utilization:
Total Total -------Synchronous I/O----- ------Asynchronous I/O------
Job or CPU Sync Async PAG DB DB Non-DB Non-DB DB DB Non-DB Non-DB
Task User Number Pty Util I/O I/O Fault Read Write Read Write Read Write Read Write
---------- ---------- ------ --- ---- ----- ----- ----- ----- ----- ------ ------ ----- ----- ------ ------
QCQSARTR QSVCCS ð93261 35 17.9 45ð 21 7 35 ð 387 28 4 2 ð 15
QCQRCVDS QSVMSS ð93254 2ð 2.7 156 128 ð ð ð 152 4 ð ð ð 128
QPADEVððð3 PITA ð93215 2ð 2.3 291 43 3 ð ð 186 1ð5 ð 1 ð 42
QPADEVððð4 ALDO ð93219 2ð 1.8 157 29 2 ð ð 127 3ð ð ð ð 29
CFINT1 ð 1.2 ð ð ð ð ð ð ð ð ð ð ð
WRKSYSACT RAMON ð93253 1 .7 45 2ð ð ð 3ð 1 14 ð 2ð ð ð
SMPOððð2 ð .2 24 ð ð ð ð ð 24 ð ð ð ð
QSYSWRK QSYS ð9313ð ð .1 24 ð ð ð ð 4 2ð ð ð ð ð
QTGTELNETS QTCP ð93172 2ð .1 ð ð ð ð ð ð ð ð ð ð ð
QPADEVððð5 MUTH ð932ð5 2ð .ð 9 ð ð 2 ð 7 ð ð ð ð ð
Order by Total I/O:
Total Total -------Synchronous I/O----- ------Asynchronous I/O------
Job or CPU Sync Async PAG DB DB Non-DB Non-DB DB DB Non-DB Non-DB
Task User Number Pty Util I/O I/O Fault Read Write Read Write Read Write Read Write
---------- ---------- ------ --- ---- ----- ----- ----- ----- ----- ------ ------ ----- ----- ------ ------
QDCPOBJ2 QSYS ð93115 6ð .ð ð ð ð ð ð ð ð ð ð ð ð
QDCPOBJ1 QSYS ð93114 6ð .ð ð ð ð ð ð ð ð ð ð ð ð
SCPF QSYS ðððððð 4ð .ð 17 5 ð 2 6 7 2 ð 1 ð 4
LCTRS ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMASPTASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMPOððð3 ð .ð 73 ð ð ð 1 ð 72 ð ð ð ð
SMPOððð2 ð .2 24 ð ð ð ð ð 24 ð ð ð ð
SMPOððð1 ð .ð 49 ð ð ð 1 ð 48 ð ð ð ð
QPADEVðð15 RAMON ð93231 2ð .ð ð ð ð ð ð ð ð ð ð ð ð
QTGTELNETS QTCP ð93172 2ð .1 ð ð ð ð ð ð ð ð ð ð ð

Figure 6-5. Sample Summary Activity Report

The header portion of this report contains the following information:

Report title The title of the report.

Current date and time The date and time when this report was printed.
Report page number The page number currently being printed.
User-selected report title The title specified by the user on the TITLE parameter
of the Print Activity Report command.
Member The name of the member in QAITMON that contained the performance
data.
Library The library where QAITMON was located.
Report type The type of report, either summary or detail, being printed.
Version The version of the Performance Tools licensed program that collected
the data.

6-10 Performance Tools V4R2

Release The release level of the Performance Tools licensed program that col-
lected the data.
Period start date and time The start date and time of the period during which the
performance statistics being printed were collected.
Period end date and time The end date and time of the period during which the
performance statistics being printed were collected.

The columns in the summary activity report are:

Job or Task The name of the job or task for which the performance statistics are
being printed.
User The user profile associated with the job.
Number The number assigned to the job.
Pty The priority at which the job or task was running when the performance
statistics were first collected.
CPU Util The percentage of the specified period during which the processing unit
was used by the job or task. For a multiple-processor system, this is the
total utilization divided by the number of processors.
Total Sync I/O The total number of synchronous physical disk I/O operations per-
formed by the job or task during the specified period. This value is the
sum of the synchronous database/nondatabase reads and writes.
Total Async I/O The total number of asynchronous physical disk I/O operations
started by the job or task during the specified period. This value is the
sum of the asynchronous database/nondatabase reads and writes.
PAG Fault The number of process access group (PAG) faults caused by the job or
task during the specified period.
Synchronous I/O DB Read The number of synchronous database read operations
performed by the job or task during the specified period.
Synchronous I/O DB Write The number of synchronous database write operations
performed by the job or task during the specified period.
Synchronous I/O Non-DB Read The number of synchronous nondatabase read
operations performed by the job or task during the specified period.
Synchronous I/O Non-DB Write The number of synchronous nondatabase write
operations performed by the job or task during the specified period.
Asynchronous I/O DB Read The number of asynchronous database read oper-
ations started by the job or task during the specified period.
Asynchronous I/O DB Write The number of asynchronous database write oper-
ations started by the job or task during the specified period.
Asynchronous I/O Non-DB Read The number of asynchronous nondatabase read
operations started by the job or task during the specified period.
Asynchronous I/O Non-DB Write The number of asynchronous nondatabase write
operations started by the job or task during the specified period.
Note: The asynchronous I/O operations are performed by system asynchronous
I/O tasks.

Chapter 6. System Activity 6-11

Detail Activity Report
For each interval available in the specified period, the Detail Activity Report lists the
performance statistics for the number of entries specified by the Number of Jobs
(NBRJOBS) parameter. The entries are ordered according to the Sequence (SEQ)
parameter.

The following represents a sample Detail Activity Report.

System Activity Report 5/3ð/98 11:ð6:38
Page 1
Member . . . . : QAITMON Report Type . . . . : DETAIL Version . . . . : 4 Started . . . . : ð5/3ð/98 1ð:56:22
Library . . . . : QPFRDATA Sequence . . . . . : CPU Release . . . . : 2.ð Stopped . . . . : ð5/3ð/98 1ð:57:ð9
Time . . . . : 1ð:56:22 Total CPU Utilization . . . . . . . . . . . : .ð
Total Total -------Synchronous I/O----- ------Asynchronous I/O------
Job or CPU Sync Async PAG DB DB Non-DB Non-DB DB DB Non-DB Non-DB
Task User Number Pty Util I/O I/O Fault Read Write Read Write Read Write Read Write
---------- ---------- ------ --- ---- ----- ----- ----- ----- ----- ------ ------ ----- ----- ------ ------
WRKSYSACT RAMON ð93253 1 .2 ð ð ð ð ð ð ð ð ð ð ð
RMSRVCTKLO ð .ð ð ð ð ð ð ð ð ð ð ð ð
LIDMGR-TAS K--AHT ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMASPTASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMASPAGENT TASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMCFGVALID ATER ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMCFGUPDAT ER ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMSLSSERVI CETASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
IOPI-HRI-P ERS-IO ð .ð ð ð ð ð ð ð ð ð ð ð ð
XMERRLOGER ð .ð ð ð ð ð ð ð ð ð ð ð ð
Time . . . . : 1ð:56:27 Total CPU Utilization . . . . . . . . . . . : .ð
Total Total -------Synchronous I/O----- ------Asynchronous I/O------
Job or CPU Sync Async PAG DB DB Non-DB Non-DB DB DB Non-DB Non-DB
Task User Number Pty Util I/O I/O Fault Read Write Read Write Read Write Read Write
---------- ---------- ------ --- ---- ----- ----- ----- ----- ----- ------ ------ ----- ----- ------ ------
WRKSYSACT RAMON ð93253 1 .7 8 2 ð ð 3 1 4 ð 2 ð ð
QPADEVððð5 MUTH ð932ð5 2ð .5 9 ð ð 2 ð 7 ð ð ð ð ð
QLZPSERV QUSER ð93239 2ð .ð ð ð ð ð ð ð ð ð ð ð ð
RMSRVCTKLO ð .ð ð ð ð ð ð ð ð ð ð ð ð
LIDMGR-TAS K--AHT ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMASPTASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMASPAGENT TASK ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMCFGVALID ATER ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMCFGUPDAT ER ð .ð ð ð ð ð ð ð ð ð ð ð ð
SMSLSSERVI CETASK ð .ð ð ð ð ð ð ð ð ð ð ð ð

Figure 6-6. Sample Detail Activity Report

The header portion of this report contains the same information as found on the
summary report, except for the Sequence field, which defines the order of the
entries listed for each interval. The value found in this field corresponds to the
value specified for the sequence (SEQ) parameter on the Print Activity Report
command.

The body of the Detail Activity Report contains the same columns of information
found on the summary report. There are, however, two additional fields associated
with the statistics for each interval:
Time The end time for the collection interval.
Total CPU Utilization The processing unit use for the entire system during the col-
lection interval.
For a multiple-processor system, this is the total utilization divided by the number of
processors.

6-12 Performance Tools V4R2

 Chapter 7. Performance Reports—Manager Feature
Performance reports provide a way for you to effectively research areas of the
system that are causing performance problems. After you have collected perform-
ance data over time, different reports offer you ways to see how and where system
resources are used. Performance reports can direct you to specific application pro-
grams, users, or inefficient workloads that are causing lower overall response
times.
| Note: While the performance reports have no exact restrictions on the amount of
| performance data that they can process, in general, it is recommended that
| a collection be no longer than one week in length. In fact, the optimal col-
| lection strategy is to run a new collection every day.

Several types of performance reports show data focusing on different aspects of

the system. For example, one report identifies CPU use and another identifies com-
munications use. These reports help identify various performance problems: if you
get complaints that the user sign-ons are taking longer than they should, you could
use a Transaction Report to find out how many CPU seconds are used by the
sign-on. You could then use a Transition Report to more closely identify how those
CPU seconds are used.

A Performance Report
| System Report 3/2ð/98 14:ð6:27
| Disk Utilization Page ððð5
| Sample System Report
| Member . . . : Q98224ð9ð3 Model/Serial . : 4ðð-2133/1ð-3594G Main storage . . : 96.ð M Started . . . . : ð3/12/98 ð9:ð4:28
| Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð3/12/98 ð9:ð4:35
| Unit Size IOP IOP Dsk CPU ASP --Percent-- Op Per K Per - Average Time Per I/O --
| Unit Name Type (M) Util Name Util ID Full Util Second I/O Service Wait Response
| ---- ---------- ---- ------- ---- ------------ ------- --- ---- ---- --------- --------- ------- ------ --------
| ððð1 DDðð1 66ð6 1,967 47.3 CMBð1 .ð ð1 76.9 33.3 3ð.2ð 4.8 .ð11ð .ðð37 .ð147
| ððð2 DDðð2 66ð6 1,967 47.3 CMBð1 .ð ð1 74.2 44.4 31.25 7.3 .ð142 .ðð35 .ð177
| ððð3 DDðð3 66ð6 1,967 47.3 CMBð1 .ð ð1 74.ð .ð 25.4ð 4.3 .ðððð .ðððð .ðððð
| ððð4 DDðð4 66ð6 1,967 47.3 CMBð1 .ð ð1 74.2 33.3 29.2ð 6.7 .ð114 .ð19ð .ð3ð4
| Total 7,868
| Average 74.9 27.7 28.89 5.8 .ðð95 .ðð68 .ð163
| Unit -- Disk arm identifier
| Unit Name -- Disk arm resource name
| Type -- Type of disk
| Size (M) -- Disk space capacity in millions of bytes
| IOP Util -- Percentage of utilization for each Input/Output Processor
| IOP Name -- Input/Output Processor resource name
| Dsk CPU Util -- Percentage of Disk Processor Utilization
| ASP ID -- Auxiliary Storage Pool ID
| Percent Full -- Percentage of disk space capacity in use
| Percent Util -- Average disk operation utilization (busy)
| Op per Second -- Average number of disk operations per second
| K Per I/O -- Average number of kilobytes (1ð24) transferred per disk operation
| Average Service Time -- Average disk service time per I/O operation
| Average Wait Time -- Average disk wait time per I/O operation
| Average Response Time -- Average disk response time per I/O operation

Figure 7-1. A Performance Report

Note: This report is only provided as an example of the layout of a report. See
each specific report example for current report details.

 Copyright IBM Corp. 1998 7-1

Performance Report Header
Each report, regardless of the type or section, contains information in the header of
the report that identifies characteristics of the data:
Report title The first line identifies the type of performance report. The second line
identifies the section of the report.
Current date and time Is the date and time the report was printed.
Report page number Identifies the page of the report.
User-selected report title Is the name assigned to the report by a user.
Data member name Is the performance data member used in the report. This
name corresponds to the name used on the MBR parameter of the Start
Performance Monitor (STRPFRMON) command.
Library name Is the library where the performance data used in the report is
located.
Model number 4ðð is the model number and 2133 is the feature code of the
AS/400 system on which the performance data was collected for this
report.
Serial number 1ð-3594G is the serial number of the system unit in this example.
The serial number can be 10 characters.
System name ABSYSTEM is the name of the system on which the performance data
was collected for this report.
Main storage size Is the size of the main storage on the AS/400 system on which
the performance data was collected in this example.
OS/400 version and release level x/ x.ð indicates which version and release
level that the AS/400 system was running.
Data collection start date and time Is the date and time the performance monitor
started collecting performance data in this example. Depending on
whether or not you select specific intervals or a specific starting time,
you could see the following:
Ÿ If you specify no intervals at which to run the report, the start date
and time is the date and time at which the data was collected.
Ÿ If you specify specific intervals at which to run the report, the start
date and time is the date and time at which the data was collected.
Note: For the system report only, you should consult the Report
Selection Criteria section to find out which intervals were
selected.
Data collection stop date and time Is the date and time the performance monitor
stopped collecting performance data in this example. Depending on
whether or not you select specific intervals or a specific ending time, you
could see the following:
Ÿ If you specify no intervals at which to run the report, the stop date
and time is the date and time at which the data was collected.
Ÿ If you specify specific intervals at which to run the report, the stop
date and time is the date and time at which the data was collected.

7-2 Performance Tools V4R2

 Note: For the system report only, you should consult the Report
Selection Criteria section to find out which intervals were
selected.
Column headings Each report also has several columns that make up the informa-
tion of the report. Some are specific to a particular report and others are
consistent between reports. In this example, IOP Util is one of the
column headings. For short descriptions of these columns, see “Per-
formance Report Columns” on page 7-102.

Available Performance Reports

System Report This report has the following sections:
Ÿ Workload
Ÿ Resource Utilization
Ÿ Resource Utilization Expansion
Ÿ Storage Pool Utilization
Ÿ Disk Utilization
Ÿ Communications Summary
Ÿ Report Selection Criteria
Component Report This report has the following sections:
Ÿ Component Interval Activity
Ÿ Job Workload Activity
Ÿ Storage Pool Activity
Ÿ Disk Activity
Ÿ IOP Utilizations
Ÿ Local Work Stations–Response Time Buckets
Ÿ Remote Work Station–Response Time Buckets
Ÿ Exception Occurrence Summary and Interval Counts
Ÿ Database Journaling Summary
Ÿ Report Selection Criteria
Transaction Report This report has the following sections:
Ÿ Job Summary Report
– Job Summary
– System Summary Data
– Distribution of Transactions by CPU/Transaction
– Transaction Significance
– Interactive Transactions by 5-Minute Intervals
– Interactive Throughput by 5-Minute Intervals
– Interactive CPU
– Interactive Response Time by 5-Minute Intervals

Chapter 7. Performance Reports—Manager Feature 7-3

 – Scatter diagram
– Interactive Program Transaction Statistics
– Summary of Seize/Lock Conflict by Object
– Report Selection Criteria
The default for the OPTION parameter on the PRTTNSRPT
command is *SS (special system information). If you leave this
default, the following special summary sections print:
– Priority-Jobtype-Pool Statistics
– Job Statistics
– Interactive Program Statistics
– Individual Transaction Statistics
– Longest Seize/Lock Conflicts
– Longest Holders of Seize/Lock Conflicts
– Batch Job Analysis
| – Concurrent Batch Job Statistics
Ÿ Transaction Report
Ÿ Transition Report
Lock Report
Job Interval Report This report has the following sections:
Ÿ Interactive Job Summary
Ÿ Noninteractive Job Summary
Ÿ Interactive Job Detail
Ÿ Noninteractive Job Detail
Ÿ Report Selection Criteria
Pool Interval Report This report has the following sections:
Ÿ Subsystem Activity
Ÿ Pool Activity
Ÿ Report Selection Criteria
Resource Interval Report This report has the following sections:
Ÿ Disk Utilization Summary
Ÿ Disk Utilization Detail
Ÿ Communications Line Detail
– SDLC Protocol
– X.25 Protocol
– TRLAN Protocol
– ELAN Protocol
– ASYNC Protocol
– BSC Protocol

7-4 Performance Tools V4R2

 – ISDN Network Interface
– Network Interface Maintenance Channel for ISDN
– IDLC Protocol
– DDI Protocol
– Frame Relay Protocol
Ÿ IOP Utilizations
Ÿ Local Work Station Response Times
Ÿ Remote Work Station Response Times
Batch Job Trace Report This report has the following sections:
Ÿ Job Summary

Which Report Do I Want?

The “System Report” on page 7-17 supplies you with an overview of how the
system is operating. It contains summary information on the workload, resource
use, storage pool utilization, disk utilization, and communications. This is a good
report to run and print often, giving you a general idea of system use. You can print
selected sections of this report.

The “Component Report” on page 7-25 supplies you with information about the
same components of system performance as a System Report, but at a greater
level of detail. This report helps you find which jobs are consuming high amounts of
system resources, such as CPU, disk, and so on.

The “Job Interval Report” on page 7-66, “Pool Interval Report” on page 7-72, and
“Resource Interval Report” on page 7-76 provide the same information as the
System Report and Component Report do, but on an interval-by-interval basis.

The “Lock Report” on page 7-63 provides information about lock and seize conflicts
during system operation.

The “Batch Job Trace Report” on page 7-89 shows the progression of different job
types (for example, batch jobs) traced through time.

The “Transaction Report” on page 7-36 provides detailed information about the
transactions that occurred during the performance data collection.

Printing Performance Reports

You can print reports using the performance data collected with the Start Perform-
ance Monitor (STRPFRMON) command.

Use the following commands to print reports:

Ÿ Print System Report (PRTSYSRPT)
Ÿ Print Component Report (PRTCPTRPT)
Ÿ Print Transaction Report (PRTTNSRPT)
Ÿ Print Lock Report (PRTLCKRPT)

Chapter 7. Performance Reports—Manager Feature 7-5

 Ÿ Print Job Report (PRTJOBRPT)
Ÿ Print Pool Report (PRTPOLRPT)
Ÿ Print Resource Report (PRTRSCRPT)
Ÿ Print Trace Report (PRTTRCRPT)

Using Menus to Print Performance Reports

1. To start Performance Tools, use the Start Performance Tools (STRPFRT)
command or type go perform on a command line.
2. To print selected information from the data you collect, choose option 3 (Print
performance report) on the IBM Performance Tools menu. The Print Perform-
ance Report display appears.
The member name, a text description, and the date and time you collected
each set of performance data appear on this display. If you cannot find the data
you want to print in a report, use the appropriate function key to sort the sets of
performance data. You can sort them by member name, text description, or by
the date and time the member was created.
3. When you find the performance data, select the type of report you want by
typing one of the following options that corresponds to the type of report:
Option Description
1 System Report
2 Component Report
3 Transaction Report
4 Lock Report
5 Job Report
6 Pool Report
7 Resource Report
8 Batch Job Trace Report

à Print Performance Report

ð
Library . . . . . . QPFRDATA

Type option, press Enter.

1=System report 2=Component report 3=Transaction report 4=Lock report
5=Job report 6=Pool report 7=Resource report 8=Batch job trace report

Option Member Text Date Time

1 TUEDTA ð3/17/98 12:2ð:29
_ DLTTEST18 ð3/17/98 1ð:1ð:2ð
_ DLTTEST17 ð3/17/98 1ð:1ð:ð2
_ DLTTEST16 ð3/17/98 1ð:ð9:42
_ DLTTEST15 ð3/17/98 1ð:ð9:32
_ DLTTEST14 ð3/17/98 1ð:ð9:22
_ DLTTEST13 ð3/17/98 1ð:ð9:ð4
_ DLTTEST11 ð3/17/98 1ð:ð8:56
_ DLTTEST1ð ð3/17/98 1ð:ð8:49
_ DLTTEST9 ð3/17/98 1ð:ð8:44
_ DLTTEST8 ð3/17/98 1ð:ð8:35
More...
F3=Exit F5=Refresh F11=Work with your spooled output files F12=Cancel
F15=Sort by member F16=Sort by text

7-6 Performance Tools V4R2

 Note: When the monitor is running and using one of the members shown in
the Print Performance Report display, this member may appear with
blank Date and Time fields until the first interval is collected.
4. To print a Transaction Report, Lock Report, or Batch Job Trace Report, press
the Enter key and the parameters for the corresponding command appear.
The Select Sections for Report display appears when you select to print the
Job Report (PRTJOBPRT).

à Select Sections for Report

ð
Member . . . . . . . . . . . : Q961ð3ð917

Type options, press Enter. Press F6 to print entire report.

1=Select

Option Section
Workload
Resource Utilization
Resource Utilization Expansion
Storage Pool Utilization
Disk Utilization
Communication Summary

Bottom
F3=Exit F6=Print entire report F12=Cancel

The Select Categories for Report display appears when you select to print one
of the following reports:
Ÿ System Report
Note: The Select Sections for Report display is shown first, followed by
the Select Categories for Report display.
Ÿ Component Report
Ÿ Job Report
Ÿ Pool Report
Ÿ Resource Report

Chapter 7. Performance Reports—Manager Feature 7-7

 à Select Categories for Report
ð
Member . . . . . . . . . . . : TUEDTA

Type options, press Enter. Press F6 to print entire report.

1=Select

Option Category
_ Time interval
_ Job
_ User ID
_ Subsystem
_ Pool
_ Communications line
_ Control unit
_ Functional area

Bottom
F3=Exit F6=Print entire report F12=Cancel

The name of the performance data member you chose on the Print Perform-
ance Report display appears at the top of the Select Categories for Report
display.
5. To include all categories of information in your report, Press F6. To limit the
amount of information in the report, type a 1 in the Option column next to those
categories of information for which you want performance data. Press the Enter
key.
For example, if you choose the Time interval option, the Select Time Intervals
display appears. This display shows an interactive view of some of the key per-
formance parameters of the data collected. The member name you typed on
the Print Performance Reports display appears in the Performance data field.
The intervals you defined to collect the performance data appear.

à Select Time Intervals

ð
Library . . . . . : QPFRDATA Performance data . . . . . : TUEDTA

Type options, press Enter.

1=Select
CPU High Pool
Transaction -Utilization- --Util-- -Fault/Sec-
Opt Date Time Count Resp Tot Inter Bch Dsk Unit Mch User ID Excp
_ ð3/17 12:39 33 1.5 3 2 ð 2 ðð17 ð 1 ð3 77
_ ð3/17 12:44 26 .9 1 1 ð 1 ððð2 ð ð ð3 7
_ ð3/17 12:49 2ð .2 1 ð ð 1 ððð9 ð ð ð3 7

F3=Exit F5=Refresh F12=Cancel

F13=Sort (date/time) F14=Sort (count) F24=More keys

6. Use the Select Time Intervals display to choose specific time intervals from the
performance data to produce a report. You should select specific time intervals
to help you manage the volume of data associated with the performance meas-
urement. The Select Time Intervals display allows you to interactively select the

7-8 Performance Tools V4R2

 time intervals of interest. This selection reduces the amount of processing
required to produce the requested report, and also reduces the size of the
resulting report.
To select time intervals to print on your report, type a 1 in the Opt column next
to the appropriate intervals. When you select multiple intervals, they are com-
bined to create a single report.
If it is difficult to find the time interval you need, you can sort the intervals differ-
ently before making your selection. You can choose to sort the time intervals in
any of the following ways:
Ÿ Date/time
Ÿ Transaction count
Ÿ Response time
Ÿ Total processing unit time
Ÿ Interactive processing unit time
Ÿ Batch processing unit time
Ÿ Disk utilization
Ÿ Machine pool faults
Ÿ User pool faults
Ÿ Exceptions
If you choose to print the report with only certain categories of information, a
display appears for each category. For example, if you choose Pool, the Select
or Omit Pools display appears.
7. Use the Select or Omit Pools display to select pools to include or omit from
your report. To use this display, type the number for the pools you want to
select or omit. If you do not know the pool numbers to select, press F4
(Prompt) to see a list of pools that were active during the collection of perform-
ance data.

à Select or Omit Pools

ð
Member . . . . . . . . . . . : TUEDTA

Type options, press Enter.

1=Select 2=Omit

Option Pool Text

_ ð1 Machine pool
_ ð2 Base pool
_ ð3
_ ð4
Bottom

F12=Cancel

Type a 1 in the Option column next to the items you want to include in your
report. Or type a 2 if you want certain items omitted from your report.

Chapter 7. Performance Reports—Manager Feature 7-9

 Note: You cannot use both the Select and Omit options at the same time. You
must indicate either the items to select or the items to omit.
To include all the items in the report, leave the Option column blank, and press
the Enter key.
For each category you choose on the Select Categories for Report display, you
must complete one of the following corresponding displays:
Ÿ Select Time Intervals
Ÿ Select or Omit Pools
Ÿ Select or Omit Jobs
Ÿ Select or Omit User IDs
Ÿ Select or Omit Subsystems
Ÿ Select or Omit Communications Lines
Ÿ Select or Omit Control Units
Ÿ Select or Omit Functional Areas
When you choose the information you want to appear on your report from the
options shown on these displays, the Select Report Options display appears.
Following is an example of the display that appears if you did not use the
Select Time Intervals display to choose any time intervals:

à Specify Report Options

ð
Type choices, press Enter.

Report title . . . . New data entry in production_________________

Start:
Day . . . . . . . \FIRST__ \FIRST, MM/DD/YY
Time . . . . . . . \FIRST__ \FIRST, HH:MM:SS

Stop:
Day . . . . . . . \LAST___ \LAST, MM/DD/YY
Time . . . . . . . \LAST___ \LAST, HH:MM:SS

Omit system tasks . \YES \YES, \NO

8. Specify the start and stop date and time. If you do not specify the start and
stop date and time, the report includes data from the first (or only) date that
data was collected, to the last (or only) date that data was collected. You may
also type a report title in the Report title field and specify whether or not you
want your report to include the system tasks. Press the Enter key to process
and print your report.
Note: The Omit system tasks field appears only if you requested printing of a
System Report.
If you made use of the Select Time Intervals display, the following version of
the Select Report Options display appears instead:

7-10 Performance Tools V4R2

 à Specify Report Options
ð
Type choices, press Enter.

Report title . . . . _________________________________________________

Omit system tasks . \YES \YES, \NO

If you so choose, type a report title in the Report title field. Also, you can
specify whether or not to include the system tasks in your report. Press the
Enter key to process and print your report.

Using Defaults to Print Performance Reports

The example in this section describes how to collect performance data and print a
System Report using the default values on the displays shown.
Note: These values are the defaults for the STRPFRMON and PRTSYSRPT com-
mands.

Use the Start Performance Tools (STRPFRT) command to start Performance Tools.

à ð
PERFORM IBM Performance Tools for AS/4ðð
System: RCHXXXXX
Select one of the following:

1. Select type of status

2. Collect performance data
3. Print performance report
4. Capacity planning/modeling
5. Performance utilities
6. Configure and manage tools
7. Display performance data
8. System activity
9. Performance graphics
1ð. Advisor

7ð. Related commands

Make the choices shown on the following displays to collect performance data and
print a System Report using the system defaults.
1. To start collecting performance data, choose option 2 (Collect performance
data) on the IBM Performance Tools menu, and press the Enter key. The
Collect Performance Data display appears.

à Collect Performance Data

ð
Select one of the following:

1. Start collecting data

2. Stop collecting data
3. Work with performance collection

2. Choose option 1 (Start collecting data).

3. Press the Enter key. The Start Collecting Data display appears.

Chapter 7. Performance Reports—Manager Feature 7-11

 à Start Collecting Data
ð
Select one of the following:

1. Collect data with defaults

2. Collect data with menus
3. Collect data with command

4. Choose option 1 (Collect data with defaults).

5. Press the Enter key. The Collect Data with Defaults display appears.

à ð
Collect Data with Defaults

Type choices, press Enter.

Member . . . . . . . . Thursdata_ Name

Library . . . . . . QPFRDATA__ Name

Text . . . . . . . . . System performance for Thursday

Time duration:
Hours . . . . . . . 9 ð-999
Minutes . . . . . . ð ð-6ð

6. Type an appropriate name for the performance data in the Member field. In this
example, the performance data is called Thursdata.
7. Type an appropriate description for the performance data you want to collect in
the Text field. In this example, the description is System performance for
Thursday.
8. Type how long you want to collect performance data in the Time duration fields.
In this example, 9 is typed in the Hours field so performance data is collected
for 9 hours.
9. Press the Enter key to start the data collection process.
10. Press F3 (Exit).
Wait for the data collection process to end before you complete the next steps.
In this example, you would wait for 9 hours before you would continue.
11. Once the data collection process is complete, use the STRPFRT command to
start Performance Tools.
12. Choose option 3 (Print performance report) on the IBM Performance Tools
menu. The Print Performance Report display appears.

à Print Performance Report

ð
Library . . . . . . QPFRDATA

Type option, press Enter.

1=System report 2=Component report 3=Transaction report 4=Lock report
5=Job report 6=Pool report 7=Resource report 8=Batch job trace report

Option Member Text Date Time

1 THURSDATA System performance for Thursday ð3/19/98 12:2ð:29
DLTTEST18 ð3/19/98 1ð:1ð:2ð
DLTTEST17 ð3/19/98 1ð:1ð:ð2

7-12 Performance Tools V4R2

13. Type a 1 in the Option column, next to the performance member called
THURSDATA, to indicate that you want to print a System Report from this perform-
ance data.
14. Press the Enter key. The Select Categories for Report display appears.

à Select Categories for Report

ð
Member . . . . . . . . . . . : THURSDATA

Type options, press Enter. Press F6 to print entire report.

1=Select

Option Category
_ Time interval
_ Job
_ User ID
_ Subsystem
_ Pool
_ Communications line
_ Control unit
_ Functional area

15. Press F6 to indicate that you do not want to restrict the categories for this
report. The Specify Report Options display appears.

à Specify Report Options

ð
Type choices, press Enter.

Report title . . . . System performance for Thursday

Start:
Day . . . . . . . \FIRST \FIRST, MM/DD/YY
Time . . . . . . . \FIRST \FIRST, HH:MM:SS

Stop:
Day . . . . . . . \LAST \LAST, MM/DD/YY
Time . . . . . . . \LAST \LAST, HH:MM:SS

16. Press the Enter key to submit a batch request to print a System Report for the
entire data collection period.
17. Press F3 (Exit) to go to the IBM Performance Tools menu.
The batch request you submit takes a period of time to complete, depending on
the amount of data collected. Use the Work with Submitted Job (WRKSBMJOB)
command to check the status of the request.
After the System Report has been produced, you can view it online and direct it
to an active writer by following steps 18 through 21.

Chapter 7. Performance Reports—Manager Feature 7-13

 à PERFORM IBM Performance Tools for AS/4ðð
ð
System: ABSYSTEM
Select one of the following:

1. Select type of status

2. Collect performance data
3. Print performance report
4. Capacity planning/modeling
5. Performance utilities
6. Configure and manage tools
7. Display performance data
8. System activity
9. Performance graphics
1ð. Advisor

7ð. Related commands

Selection or command
===> WRKSPLF____________________________________________________________
________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant
F16=System main menu

á ñ
18. Type WRKSPLF on the command entry line, and press the Enter key.
The Work with All Spooled Files display appears.

à Work with All Spooled Files

ð
Type options, press Enter.
2=Change 3=Hold 4=Delete 5=Display 6=Release 8=Attributes

Device or Total Cur

Opt File User Queue User Data Sts Pages Page Copy
2 QPPTSYSR USERID QSYSPRT RDY 7 1

Parameters for option 2 or command

===> OUTQ(outqname)_____________________________________________________
F3=Exit F1ð=View 3 F11=View 2 F12=Cancel F24=More keys

á ñ
On this display you could choose option 5, for example, to view the System
Report online.
19. For this example, type a 2 under the Opt column to change the output queue
for the System Report (the QPPTSYSR file). In this example, you might want to
move the report to an output queue that has an active writer, so the report
prints on the device the writer is associated with.
20. Type the new output queue name. To do this, type OUTQ(outqname) on the
command entry line.
21. Press the Enter key. The System Report prints when a device is available.

7-14 Performance Tools V4R2

Why Performance Reports May Seem Inconsistent
Performance Tools pulls data from the various performance monitor database files
to create the performance reports. As a result, some values in the report's columns
are inconsistent between reports where it seems like they should match.

For example, the Communications Summary report (System Report) uses the
QAPMJOBS file. QAPMJOBS file records batch use for jobs that are not related to
communications. As a result, batch use of a line or TCP use does not show up in
the QAPMJOBS file. Because the QAPMJOBS file only shows transactions for jobs,
and the communications line connected to the job is classified as interactive, no
batch use for communications is recorded by the QAPMJOBS file.
Note: The Communications Summary (System Report) only shows interactive
results. Besides this section only takes information from the records which
have data in the line description, communications line name, or in the sec-
ondary line description, pass-through and emulation (only "virtual" type con-
nection).

Another example is the IOP Utilizations Component report section, which uses the
QAPMCIOP file. This file has field values for the idle loop count and the idle loop
time. These values make up the data that is used to calculate the IOP utilization
value that shows up in this report. The IOP utilization result is just the percentage
of CPU used in the IOP. When the communication IOP utilization shows a value
different than 0, it does not necessarily mean that the IOP is doing any data trans-
fers, it may just be overhead of an active line.

Another confusing example is how Client Access transactions show up on the

System report. Client Access jobs show up in the interactive category, but some
Client Access functions show up in batch or evoke categories. In addition, lots of
Client Access activity never become a real job. This can have some effect on trans-
action counting. For example, the Interactive Workload System report section
shows the different job types (Interactive, Client Access, DDM Server, Passthru,
and so on). Each of these job types has a column with the corresponding total of
transactions for each one. This report pulls data from the QAPMSYS file, where the
performance monitor assigns different classifications of data for the different job
types. The performance monitor code checks for certain attributes and flags on the
system to determine which types they are. For example, Client Access jobs are
identified by a flag. The flag is turned on when a Client Access application sets a
Client Access bit in the work control block (WCB). The performance monitor then
recognizes this job as Client Access and classifies it as such.

As a result, the number of transactions that show up on the Interactive Workload

section differs from the number of transactions that show up on the Communi-
cations Summary System report section. The Communications Summary System
report section only shows the number of interactive transactions on the communi-
cation line over a period of time (which includes Client Access transactions). The
transactions that are unrelated to the communications line do not show up in this
section of the report.

Performance Tools reports show the data based on the contents of the perform-
ance monitor database files. In some cases, this causes slight inconsistencies
between reports.

Chapter 7. Performance Reports—Manager Feature 7-15

| Table 7-1 on page 7-16 identifies the type of workload that is running on the
| system and shows how the System Report (SYS), the Component Report (CPT),
| and the Transaction Report (TNS) report the job type for the QAPMJOBS database
| file.

| The abbreviations for the field value headings include the following:
| Ÿ JBTYPE - job type
| Ÿ JBSTYP - job subtype
| Ÿ JBPTTF - target pass-through flag
| Ÿ JBPTSF - source pass-through flag
| Ÿ JBEAF - emulation active flag
| Ÿ JBPCSF - Client Access flag
| Ÿ JBDDMF - target DDM job flag

| The Desc column identifies the type of workload that is running on the system. This
| column contains a number that is associated with the following descriptions:
| Ÿ 1 - 5250 twinaxial data link control, remote workstation support, or 3270 remote
| attach
| Ÿ 2 - APPC 5250 emulation (Client Access)
| Ÿ 3 - Target APPC display station pass-through
| Ÿ 4 - Target TELNET 5250
| Ÿ 5 - Source pass-through
| Ÿ 6 - Target distributed data management (DDM)
| Ÿ 7 - APPC router
| Ÿ 8 - Host server ("Client Access"), pre-started job
| Ÿ 9 - APPC, batch evoke
| Ÿ 10 - Normal batch job
| Ÿ 11 - Auto start job
| Ÿ 12 - Subsystem monitor
| Ÿ 13 - Spool writer
| Ÿ 14 - Spool print driver
| Ÿ 15 - Other system jobs
| Ÿ 16 - typical secondary thread

| You can find the descriptions for the one-character and two-character abbreviations
| used in the table under the Typ column description at the end of this chapter.

| Table 7-1 (Page 1 of 2). Job Types

| Desc Jobtype Listed in QAPMJOBS Field Values1
| Reports
| SYS CPT TNS JBTYPE JBSTYP JBPTTF JBPTSF JBEAF JBPCSF JBDDMF
| 1 I2 I I I b 0 0 0 0 0

7-16 Performance Tools V4R2

| Table 7-1 (Page 2 of 2). Job Types
| Desc Jobtype Listed in QAPMJOBS Field Values1
| Reports
| SYS CPT TNS JBTYPE JBSTYP JBPTTF JBPTSF JBEAF JBPCSF JBDDMF
| 2 CA2,3 C I I b 1 0 0 1 0
| 3 PT2,4 P I I b 1 0 0 0 0
| 4 PT2,4 P I I b 1 0 0 0 0
| 5 NA NA NA NA b 0 1 0 0 0
| 6 DDM D BE B E 1 1 0 0 1
| server5
| 7 Batch5 E BE B E 0 0 0 1 0
| 8 Batch5 C BJ B J 0 0 0 1 0
| 9 Batch5 C B B b 0 0 0 0 0
| 10 Batch5 B B B b 0 0 0 0 0
| 11 Auto A A A b 1 1 0 0 0
| start5
| 12 System5 M M M b 0 0 0 0 0
| 13 Spool5 W W W b 0 0 0 0 0
| 14 Spool5 W WP W P 0 0 0 0 0
| 15 Batch5 S S S b 0 0 0 0 0
| 16 Batch5 B BD B D 0 0 0 0 0
| 1 A lowercase b indicates the field is blank.
| 2 Interactive workload
| 3 CA represents Client Access
| 4 PT represents pass-through
| 5 Non-interactive workload

System Report

Printing the System Report

Print the System Report using the Print System Report (PRTSYSRPT) command,
or select option 1 (System report) on the Print Performance Reports display.

What Is the System Report?

The System Report is an overview of system operation during the data collection
period. Produce and save this report periodically so you have a record of the work-
load and resource utilization (for example, how your system meets the users’ needs
and at what cost). Use the reports to see what processing trends are developing,
and to project when you might need to make application, system, or operational
changes to accommodate changing workloads.

Every System Report includes the Workload, Resource Utilization, and Resource
Utilization Expansion sections. However, the Storage Pool Utilization, Disk Utiliza-

Chapter 7. Performance Reports—Manager Feature 7-17

 tion, and Communications Summary sections are omitted when certain report cate-
gories are selected on the Select Categories for Report display. Table 7-2 on
page 7-18 shows the categories that cause these sections to be omitted.

Table 7-2. Report Categories that Cause System Report Sections to be Omitted
Report Category Report Section Omitted
Time Interval None
Job Storage Pool Utilization
Disk Utilization
Communications Summary
User ID Storage Pool Utilization
Disk Utilization
Communications Summary
Subsystem Storage Pool Utilization
Disk Utilization
Communications Summary
Pool Disk Utilization
Communications Summary
Communications line Storage Pool Utilization
Disk Utilization
Control unit Storage Pool Utilization
Disk Utilization
Functional area Storage Pool Utilization
Disk Utilization
Communications Summary

For samples of each section of the System Report, see “Sample System Reports”
on page 7-20.

For definitions of specific columns in the reports, see “Performance Report

Columns” on page 7-102.

Workload

Interactive Workload
The first part of the Workload section shows the interactive workload of the system.

See the sample report shown in Figure 7-2 on page 7-20.

Noninteractive Workload
The second part of the Workload section shows the noninteractive workload of the
system.

See the sample report shown in Figure 7-3 on page 7-21.

7-18 Performance Tools V4R2

Resource Utilization
The Resource Utilization section shows the average resource utilization per interac-
tive transaction. Use it to note changes in resource utilization from one measure-
ment period to another and to determine resource utilization trends.

Resource Utilization (First Part)

See the sample report shown in Figure 7-4 on page 7-21.

Resource Utilization (Second Part)

See the sample report shown in Figure 7-5 on page 7-21.

Resource Utilization Expansion

The Resource Utilization Expansion section gives the average resource use per
transaction by job type.

Resource Utilization Expansion (First Part)

See the sample report shown in Figure 7-6 on page 7-21.

Resource Utilization Expansion (Second Part)

The second part of the Resource Utilization Expansion section contains CPU and
I/O utilization information.

See the sample report shown in Figure 7-7 on page 7-22.

Storage Pool Utilization

Use the Storage Pool Utilization section of the System Report, along with the
guidelines provided in the Work Management book, to help you set the storage
pool size and activity level.

See the sample report shown in Figure 7-8 on page 7-22.

Disk Utilization
The Disk Utilization section of the System Report shows the utilization for each
disk.

See the sample report shown in Figure 7-9 on page 7-23.

Communications Summary
The Communications Summary section of the System Report shows the use of the
communications lines and processors.

See the sample report in Figure 7-10 on page 7-23.

Note: The line utilization in the system report shown in Figure 7-10 on page 7-23
does not correspond with the “Component Report: IOP Utilizations” for an
IOP running SDLC remote work stations. A low SDLC line utilization value
results in a high IOP utilization value due to polling. However, because the
SDLC line transfers a larger percentage of user data, an IOP polls less fre-
quently. Usually this results in an overall increase in IOP utilization. In some
cases, though, especially when the SDLC lines have a low utilization, this
results in an overall decrease in IOP utilization. Thus, a high IOP utilization
value is significant only if at least one of the attached SDLC lines is active.

Chapter 7. Performance Reports—Manager Feature 7-19

 Report Selection Criteria

Report Selection Criteria (Selected Start/End Time/Date)

The Selected Start/End Time/Date Criteria section gives the range of of time the
report must generate. If you use no SELECT Start/End Time/Date, the message No
Select Time/Date were chosen appears.

A sample report is shown in Figure 7-13 on page 7-25.

Report Selection Criteria (Date/Time Intervals)

The Selected Date/Time Intervals Criteria section gives the interval number Date
and time of the intervals selected to generate the report. If you use no SELECT
Date/Time Intervals, the message All Intervals were chosen appears.

A sample report is shown in Figure 7-14 on page 7-25.

Report Selection Criteria (Select Parameters)

The Report Selection Criteria section gives the selection values you chose to
produce the report. If you use no SELECT parameters, the message No Select
parameters were chosen appears.

A sample report is shown in Figure 7-11 on page 7-24.

Report Selection Criteria (Omit Parameters)

If you did not use OMIT parameters, the message No Omit parameters were
chosen appears.

See the sample report showing the OMIT parameters on the Report Selection Cri-
teria section of the System Report in Figure 7-12 on page 7-24.

Sample System Reports

See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Workload Section: Interactive Workload–Sample

| System Report 9/15/98 1ð:5ð:4ð
| Workload Page ððð1
| Sample System Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| QPFRADJ . . . : 2 QDYNPTYSCD . . : 1
| Interactive Workload
| Job Number Average Logical DB -------- Printer ---------- Communications MRT
| Type Transactions Response I/O Count Lines Pages I/O Count Max Time
| ------------- ---------------- -------- ----------------- ------------- ---------- -------------- --------
| Interactive 28 2.96 188 ð ð ð ð
| PassThru 2ð1 .64 481 1,ð4ð 23 ð ð
| Total 229 669 1,ð4ð 23 ð
| Average .93

Figure 7-2. Workload Section: Interactive Workload

7-20 Performance Tools V4R2

 Workload Section: Noninteractive Workload–Sample
| Non-Interactive Workload
| Job Number Logical DB --------- Printer --------- Communications CPU Per Logical
| Type Of Jobs I/O Count Lines Pages I/O Count Logical I/O I/O /Second
| ---------- --------- ---------------- ------------- ----------- -------------- ----------- -----------
| Batch 256 13,859 175,589 3,291 ð 2.ð3ð6 .1
| Spool 1 155 ð ð ð .ð19ð .ð
| AutoStart 3 12,743 ð ð ð .ðð14 .1
| Total 26ð 26,757 175,589 3,291 ð
| Average 1.ð525 .3
| Total CPU Utilization . . . . . . : 99.9

Figure 7-3. Workload Section: Noninteractive Workload

Resource Utilization (First Part)–Sample

| System Report 9/15/98 1ð:5ð:4ð
| Resource Utilization Page ððð2
| Sample System Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| -------------------- Average Per Transaction --------------------
| Sync Async
| Job Response CPU Disk Disk DB
| Type Seconds Seconds I/O I/O I/O
| -------------- -------- ------- ------- -------- ---------
| Interactive 2.96 .21 134.2 14.ð 6.7
| PassThru .64 .ð9 36.2 2.9 2.3
| Average .93 .1ð 48.2 4.2 2.9

Figure 7-4. Resource Utilization

Resource Utilization (Second Part)–Sample

| Tns Active -------------------------- Disk I/O Per Second ---------------------------
| Job CPU /Hour Jobs Per Total --------- Synchronous -------- -------- Asynchronous --------
| Type Util Rate Interval I/O DBR DBW NDBR NDBW DBR DBW NDBR NDBW
| -------------- ---- -------------- -------- ------- ----- ----- ----- ----- ----- ----- ----- -----
| Interactive .ð 1 1 .ð .ð .ð .ð .ð .ð .ð .ð .ð
| PassThru .ð 8 1 .ð .ð .ð .ð .ð .ð .ð .ð .ð
| Average .ð 9 11 .1 .ð .ð .1 .ð .ð .ð .ð .ð

Figure 7-5. Resource Utilization

Resource Utilization Expansion (First Part)–Sample

| System Report 9/15/98 1ð:5ð:4ð
| Resource Utilization Expansion Page ððð3
| Sample System Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| -------------------------------------------- Average Per Transaction --------------------------------------------
| --------------------- Physical Disk I/O ------------------------ -------- Logical --------- -- Communications --
| Job --------- Synchronous --------- -------- Asynchronous --------- ----- Data Base I/O ------ I/O
| Type DBR DBW NDBR NDBW DBR DBW NDBR NDBW Read Write Other Get Put
| -------------- ------- ------- ------- ------- ------- ------- ------- ------- ------- ------- ------- ----- -----
| Interactive 3.71 1.5ð 122.14 6.92 .75 1.71 .ð7 11.5ð 5.92 .28 .5ð .ð .ð
| PassThru .35 .33 33.19 2.33 .59 .2ð .ð7 2.ð2 2.27 .11 .ðð .ð .ð
| Average .76 .47 44.ð7 2.89 .61 .38 .ð7 3.18 2.72 .13 .ð6 .ð .ð

Figure 7-6. Resource Utilization Expansion

Chapter 7. Performance Reports—Manager Feature 7-21

 Resource Utilization Expansion (Second Part)–Sample
| Job CPU Cum ------- Disk I/O ------- ----- CPU Per I/O ------ -- DIO /Sec --
| Priority Type Util Util Sync Async Sync Async Sync Async
| -------- -------------- ---- ----- ----------- ----------- ---------- ---------- ----- -----
| ððð Batch .ð .ð 7,238 3,8ð1 .ðð52 .ðð99 .ð .ð
| System 66.6 66.6 8,262 67,3ð4 6.8427 .8399 .ð .7
| ðð5 Batch .ð 66.7 8,269 2,134 .ðð77 .ð299 .ð .ð
| ðð9 System .ð 66.7 12 ð .ððð5 .ðððð .ð .ð
| ð1ð Interactive .ð 66.7 157 ð .ððð4 .ðððð .ð .ð
| ð16 System .ð 66.7 15 ð .ððð4 .ðððð .ð .ð
| ð19 PassThru .ð 66.7 572 79 .ðð28 .ð2ð7 .ð .ð
| ð2ð Interactive .ð 66.7 3,6ð3 393 .ðð16 .ð15ð .ð .ð
| Client Access .ð 66.7 ð ð .ðððð .ðððð .ð .ð
| PassThru .ð 66.8 6,7ð8 5ð4 .ðð25 .ð341 .ð .ð
| Batch .ð 66.8 2,ð73 323 .ð377 .2425 .ð .ð
| AutoStart .ð 66.8 ð ð .ðððð .ðððð .ð .ð
| .
| .
| .
| Total 1,1ðð,764 23ð,28ð 12.9 2.7

Figure 7-7. Resource Utilization Expansion (Second Part)

Storage Pool Utilization–Sample

| System Report 9/15/98 1ð:5ð:4ð
| Storage Pool Utilization Page ððð5
| Sample System Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| ---------- Avg Per Second --------- ---- Avg Per Minute -----
| Pool Expert Size Act CPU Number Average ------ DB ------ ---- Non-DB ---- Act- Wait- Act-
| ID Cache (K) Lvl Util Tns Response Fault Pages Fault Pages Wait Inel Inel
| ---- ------- ----------- ----- ----- ----------- -------- ------- ------- ------- ------- -------- ------- -------
| \ð1 ð 58,112 ð 66.6 ð .ðð .ð .ð .6 .6 5 ð ð
| \ð2 ð 175,2ð4 44 33.2 1 2.ðð .ð .ð 2.6 8.8 241 ð ð
| ð3 ð 2,62ð 1 .ð ð .ðð .ð .ð .ð .ð ð ð ð
| \ð4 ð 26,2ð8 18 .ð 228 .92 .ð .ð .ð .1 ð ð ð
| Total 262,144 99.9 229 .ð .ð 3.3 9.6 247 ð ð
| Average
| .93
| \ The pool did not exist for all of run, or the size or activity level
| changed during run.
| Pool ID -- Pool identifier
| Expert Cache -- Method used by the system to tune the storage pool
| Size (K) -- Size of the pool in kilobytes at the time of the first sample interval
| Act Lvl -- Activity level at the time of the first sample interval
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| Number Tns -- Number of transactions processed by jobs in this pool
| Average Response -- Average transaction response time
| DB Fault -- Average number of data base faults per second
| DB Pages -- Average number of data base pages per second
| Non-DB Fault -- Average number of non-data base faults per second
| Non-DB Pages -- Average number of non-data base pages per second
| Act-Wait -- Average number of active to wait job state transitions per minute
| Wait-Inel -- Average number of wait to ineligible job state transitions per minute
| Act-Inel -- Average number of active to ineligible job state transitions per minute

Figure 7-8. Storage Pool Utilization

Disk Utilization–Sample

7-22 Performance Tools V4R2

| System Report 11/11/98 8:2ð:ð8
| Disk Utilization Page ððð6
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . : 11/11/98 ð7:58:ð1
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : 11/11/98 ð8:18:43
| Unit Size IOP IOP Dsk CPU ASP --Percent-- Op Per K Per - Average Time Per I/O --
| Unit Name Type (M) Util Util
| Name ID Full Util Second I/O Service Wait Response
| ---- ---------- ---- ------- ---- ------------ ------- --- ---- ---- --------- --------- ------- ------
| --------
| ððð1A DDð1ð 66ð6 1,967 1.8 CMBð1 .ð ð1 77.2 1.8 1.43 5.6 .ð125 .ðð14 .ð139
| ððð1B DDðð2 66ð6 1,967 1.8 CMBð1 .ð ð1 77.2 1.5 1.64 5.4 .ðð91 .ððð6 .ðð97
| ððð3 DDðð3 66ð6 1,475 6.5 SIð3 .ð ð1 74.6 .4 1.86 6.5 .ðð21 .ðððð .ðð21
| ððð4 DDðð4 66ð6 1,475 6.5 SIð3 .ð ð1 74.3 .9 2.27 5.8 .ðð39 .ððð4 .ðð43
| ððð5 DDðð5 66ð6 1,475 6.5 SIð3 .ð ð1 74.6 .8 1.65 6.ð .ðð48 .ðððð .ðð48
| ððð6 DDðð6 66ð6 1,475 6.5 SIð3 .ð ð1 74.1 .8 2.45 6.ð .ðð32 .ðððð .ðð32
| ððð7 DDð18 66ð7 3,145 6.5 SIð3 .ð ð1 74.ð 2.ð 7.9ð 6.1 .ðð25 .ððð1 .ðð26
| ððð8 DDð2ð 66ð7 3,145 6.5 SIð3 .ð ð1 74.4 .6 2.11 6.ð .ðð28 .ððð5 .ðð33
| ððð9 DDð16 66ð7 3,145 6.5 SIð3 .ð ð1 74.ð 5.ð 2ð.88 6.2 .ðð23 .ððð4 .ðð27
| ðð1ð DDð11 66ð7 3,145 6.5 SIð3 .ð ð1 74.2 .5 1.72 6.7 .ðð28 .ðððð .ðð28
| Total 22,414
| Average 74.8 1.4 4.39 6.1 .ðð31 .ððð5 .ðð36

| Unit -- Disk arm identifier

| Unit Name -- Disk arm resource name
| Type -- Type of disk
| Size (M) -- Disk space capacity in millions of bytes
| IOP Util -- Percentage of utilization for each Input/Output Processor
| IOP Name -- Input/Output Processor resource name
| Dsk CPU Util -- Percentage of Disk Processor Utilization
| ASP ID -- Auxiliary Storage Pool ID
| Percent Full -- Percentage of disk space capacity in use
| Percent Util -- Average disk operation utilization (busy)
| Op per Second -- Average number of disk operations per second
| K Per I/O -- Average number of kilobytes (1ð24) transferred per disk operation
| Average Service Time -- Average disk service time per I/O operation
| Average Wait Time -- Average disk wait time per I/O operation
| Average Response Time -- Average disk response time per I/O operation

Figure 7-9. Disk Utilization

Communications Summary–Sample
| System Report 5/19/96 12:ð3:48
| Communications Summary Page ððð6
| Sample System Report
| Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5/ð3/96 14:59:32
| Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 3/ 7.ð Stopped . . . . : ð5/ð3/96 15:ð4:31
| IOP Name/ Line Avg Max Active Number Average ------ Bytes Per Second ----
| Line Protocol Speed Util Util Devices Transactions Response Received Transmitted
| -------------------- -------- --------- ---- ---- ------- ------------ -------- ---------- -----------
| CCð1 (2619)
| TRLINLð5 TRLAN 16ððð.ð ð ð ð ð .ðð .ð .ð
| IOP Name/Line -- IOP Resource name and model number, Line ID
| Protocol -- Line protocol (SDLC, ASYNC, BSC, X25, TRLAN, ELAN, IDLC, DDI, FRLY)
| Line Speed -- Line speed (1ððð bits per second)
| (For IDLC this is the maximum over the measurement)
| Avg Util -- Average line utilization
| Max Util -- Maximum line utilization in all measurement intervals
| Active Devices -- Average number of active devices on the line
| Number Transactions -- Number of transactions
| Average Response -- Average system response (service) time (seconds)
| Bytes /Sec Received -- Average number of bytes received per second
| Bytes /Sec Transmitted -- Average number of bytes transmitted per second

Figure 7-10. Communications Summary

Chapter 7. Performance Reports—Manager Feature 7-23

System Report Selection Criteria: Select Parameters–Sample
Report Selection Criteria

Selected Start/End Time/Date:

- No Select Time/Date were chosen.

Selected Date/Time Intervals:

- All Intervals were chosen.

Select Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Select parameters were chosen.

Figure 7-11. Report Selection Criteria: Select Parameters

System Report Selection Criteria: Omit Parameters–Sample

Report Selection Criteria

Selected Start/End Time/Date:

- No Select Time/Date were chosen.

Selected Date/Time Intervals:

- All Intervals were chosen.

Omit Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Omit parameters were chosen.

Figure 7-12. Report Selection Criteria: Omit Parameters

7-24 Performance Tools V4R2

Report Selection Criteria: Selected Start/End Time/Date–Sample
Selected Start/End Time/Date:
Start Date/Time: 12/27/95 15:13:42
End Date/Time: 12/27/95 17:38:42
Selected Date/Time Intervals:

Figure 7-13. Report Selection Criteria: Start/End Time/Date

Report Selection Criteria: Date/Time Intervals–Sample

Selected Start/End Time/Date:
- No Select Time/Date were chosen.
Selected Date/Time Intervals:
Interval
Number Date Time
2 12/27 16:13
3 12/27 17:13
4 12/27 17:38

Figure 7-14. Report Selection Criteria: Date/Time Intervals

Component Report

Printing the Component Report

Use the Print Component Report (PRTCPTRPT) command, or select option 2
(Component report) on the Print Performance Reports display.

What Is the Component Report?

This series of reports, like the System Report, is produced from the sample data
you collect with the STRPFRMON command. It expands on the detail for each
component of system performance shown on the System Report. Data is shown for
each sample interval, or in some cases, for each job. Because the report can be
lengthy, you may want to use the Select Time Intervals display when requesting
this report to select only those measurement intervals of interest to you.
Note: The Component Report does not show information for Client Access
devices for the Job Workload Activity, IOP Utilizations, and the Remote
Work Stations – Response Time Buckets sections.

For samples of each section of the Component Report, see “Sample Component
Reports” on page 7-29.

For definitions of specific columns in the reports, see “Performance Report

Columns” on page 7-102.

Component Interval Activity

The Component Interval Activity section of the Component Report gives the use of
the processing unit, disks, and pools at various time intervals.

See the sample report shown in Figure 7-15 on page 7-29.

Chapter 7. Performance Reports—Manager Feature 7-25

Job Workload Activity
The Job Workload Activity section of the Component Report gives the total number
of transactions, the transactions per hour, the average response time, the number
of disk operations, the number of communications operations, the number of PAG
faults, the number of arithmetic overflows, and the number of permanent writes for
each job.

See the sample report shown in Figure 7-16 on page 7-30.

Storage Pool Activity

The Storage Pool Activity section of the Component Report gives detailed informa-
tion for each storage pool. This information includes the storage pool activity level,
as well as the number of transactions processed in each pool.

The Pool Identifier, shown at the top of the Storage Pool Activity section, specifies
the storage pool identifier (the value can be from 01 through 16). A separate
Storage Pool Activity section exists for each pool that was in use during the meas-
urement period and was selected on the PRTCPTRPT command.

See the sample report shown in Figure 7-17 on page 7-31.

Disk Activity
The Disk Activity section of the Component Report gives the average disk activity
per hour and the disk capacity for each disk.

See the sample report shown in Figure 7-18 on page 7-32.

Input/Output Processor (IOP) Utilizations

The IOP Utilizations section of the Component Report gives the input/output
processor (IOP) utilization for communications, direct access storage devices
(DASDs), multifunction (DASD, communication, and local work stations). Con-
sistent utilization, at or above the threshold value of the DASD IOP and multifunc-
tion IOP, will affect system performance and cause longer response times or less
throughput.

See the utilization guidelines and thresholds in BEST/1 Capacity Planning Tool
book for a list of threshold values.

See the sample report shown in Figure 7-19 on page 7-33.

Note: The total for the I/O processor utilization oftentimes does not match the sum
of the three columns (IOP Processor Util Comm, IOP Processor Util LWSC,
and IOP Processor Util DASD). This mismatch is caused by the utilization of
other small components, such as system time.

Local Work Stations

The Local Work Stations section of the Component Report gives the utilization of
each controller, the range of response times for each device, and the average
response time for each device.

7-26 Performance Tools V4R2

 The response time intervals are set up on the system STRPFRMON command.
The values for the response times may vary depending on the values you use in
the command.

See the sample report shown in Figure 7-20 on page 7-33.

Remote Work Stations

The Remote Work Stations section of the Component Report gives the range of
response times for each device on the displayed controllers and the average
response time for each device.

The response time intervals are set up on the system STRPFRMON command.
The values for the response times may vary depending on the values you use in
the command.
Note: This section appears only if 5494 remote work station data is included in the
data collection.

See the sample report shown in Figure 7-21 on page 7-34.

Exception Occurrence Summary and Interval Counts

The Exception Occurrence Summary and Interval Counts section of the Component
Report gives the number of exceptions that occurred and the frequency of these
exceptions.

In some cases these exception counts can be high even under normal system
operation.

See the sample report shown in Figure 7-22 on page 7-34.

Database Journaling Summary

The Database Journaling section of the Component Report provides information
about the journal activity on the system. This information is helpful in understanding
the trade-offs between the following:
Ÿ The affects of extensive journaling.
Ÿ The time required to rebuild access paths during an IPL following an abnormal
system end.
For more information on journaling, see the Backup and Recovery book.

The Database Journaling section summarizes the journaling activity resulting from
user-initiated activities and from system-managed access-path protection (SMAPP)
support. This includes the following information:
Ÿ The number of start and stop journaling operations performed.
Ÿ The number of journal entry deposits made on behalf of objects for which a
user started journaling.
Ÿ The number of journal entry deposits made on behalf of objects for which the
system started journaling.

The report contains the following fields for the number of journal deposits resulting
from system-initiated journaling:

Chapter 7. Performance Reports—Manager Feature 7-27

 Ÿ The total number of deposits.
Ÿ A subset of the total number of deposits made to journals created by a user.
The remaining journal entries were deposited to internal system journals.
Internal system journals are created and maintained by the system.

As journal entries are deposited to the journals, the system attempts to group these
entries into larger bundles to provide more efficient I/O. The number of bundles
written to user-created journals can be compared to the number of bundles written
to system-created journals. This proportion indicates how efficiently the system per-
forms I/O to the journal receivers.

When SMAPP is active on the system, the following information is also available:
Ÿ The number of exposed access paths.
Ÿ An estimate of the time in minutes required to rebuild the exposed access
paths following an abnormal system end.
Ÿ The number of adjustments made by the system to internal journal tuning
tables.
Note: The estimated rebuild time is rounded to the nearest full minute. The
estimate is available only on a system-wide basis, not by auxiliary
storage pool (ASP), even though access path recovery times may be
specified on an ASP basis.
The number of exposed access paths and their estimated rebuild exposure does
not include the following:
Ÿ Access paths that are being journaled by a user
Ÿ Access paths that were created with the *REBLD maintenance option
See the Backup and Recovery book for more SMAPP considerations.

The estimated rebuild exposure is calculated two ways:

Ÿ Current estimated system exposure
Ÿ Estimated exposure if the system was not journaling any of the exposed access
paths
These calculated values will be the same if the system access path recovery time
is set to *NONE. These values will also be the same if the system access path
recovery time is set to a time greater than the current estimated exposure.

See the sample report shown in Figure 7-23 on page 7-35.

Report Selection Criteria

The Report Selection Criteria section of the Component Report gives the selection
values you chose to produce the report.

If you did not use the SELECT parameters, the message No Select parameters
were chosen appears. If you did not use OMIT parameters, the message No Omit
parameters were chosen appears.

See the sample report shown in Figure 7-24 on page 7-35.

7-28 Performance Tools V4R2

 Sample Component Reports
See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Component Interval Activity–Sample

| Component Report 11/11/98 8:2ð:28
| Component Interval Activity Page 1
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . : 11/11/98 ð7:58:ð1
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : 11/11/98 ð8:18:43
| --- Disk I/O --- High Pool Excp
| Itv Tns Rsp DDM ---- CPU Utilization --- --- Per Second -- -Utilization- --- Faults/Sec -- per
| End /Hour /Tns I/O Total Inter Batch Sync Async Disk Unit Mch User ID Second
| ----- ----------- ----- ---------- ------ ----- ----- ------- ------- ---- ---- --- ---- -- -------
| ð8:ð3 314 2.ð3 ð 99.9 .5 98.4 4.8 2.8 2 ððð4 ð 3 ð2 7ð
| ð8:ð8 6ð9 1.94 ð 98.8 .4 97.ð 9.2 24.5 5 ððð9 ð ð ð2 81,846
| ð8:13 95 4.25 ð 98.2 .5 96.2 12.3 49.1 8 ððð9 ð 1 ð2 164,433
| ð8:18 348 .86 ð 98.4 .9 95.6 15.2 47.5 8 ððð9 ð 3 ð4 15ð,189
| ð8:18 9ð .ðð ð 99.4 .ð 98.ð 3.2 39.6 5 ððð9 ð ð ð2 11,932

| Itv End -- Interval end time (hour and minute)

| Tns /Hour -- Number of interactive transactions per hour
| Rsp /Tns -- Average interactive transaction response time in seconds
| DDM I/O -- Number of DDMs
| Total CPU Utilization -- Percentage of available CPU time used by interactive and batch jobs. This is the average
| of all processors
| Inter CPU Utilization -- Percentage of available CPU time used by interactive jobs. This is the average of all processors
| Batch CPU Utilization -- Percentage of available CPU time used by batch jobs. This is the average of all processors
| Sync Disk I/O Per Sec -- Average synchronous disk I/O operations per second
| Async Disk I/O Per Sec -- Average asynchronous disk I/O operations per second
| High Disk Utilization -- Percent of utilization of the most utilized disk arm during this interval
| High Utilization Unit -- Disk arm which had the most utilization during this interval
| Mch Pool Faults/Sec -- Average number of machine pool faults per second
| User Pool Faults/Sec -- Average number of user pool page faults per second,
| for the user pool with highest fault rate during this interval
| Pool ID -- User pool that had the highest page fault rate
| Excp -- Total number of program exceptions that occurred

Figure 7-15. Component Interval Activity

Job Workload Activity–Sample

Chapter 7. Performance Reports—Manager Feature 7-29

| Component Report 11/11/98 8:2ð:28
| Job Workload Activity Page 2
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . : 11/11/98 ð7:58:ð1
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : 11/11/98 ð8:18:43
| T P
| Job User Job y t CPU Tns ---------- Disk I/O --------- Cmn PAG Arith Perm
| Name Name Number p Pl y Util Tns /Hour Rsp Sync Async Logical I/O Fault Ovrflw Write
| ---------- ---------- ------ -- -- -- ------ ------- --------- ------- --------- --------- --------- --------- ----- ------- ------
| AM-ANSMGR L ð1 68 .ððð ð ð .ððð ð ð ð ð ð ð ð
| ASJ1LU QSNADS ð23ð84 B ð2 4ð .ððð ð ð .ððð 1 ð ð ð 1 ð ð
| CENTRAL QGATE ð23ð85 B ð2 4ð .ððð ð ð .ððð 4 ð ð ð 4 ð ð
| CFINTð1 L ð1 ðð .949 ð ð .ððð ð ð ð ð ð ð ð
| CHAINBCH QSYS ð234ð2 M ð5 ðð .ððð ð ð .ððð ð ð ð ð ð ð ð
| CHAINUPCW QPGMR ð234ð3 B ð5 5ð 31.698 ð ð .ððð 43ð2 192ð8 37ð73ð ð ð 2ð8385 11ð91
| CHAINUPCW QPGMR ð234ð4 B ð5 5ð 3ð.857 ð ð .ððð 43ð4 19ð12 357318 ð ð 1995ð3 11247
| DBDOU L ð1 2ð .ððð ð ð .ððð 5 ð ð ð ð ð ð
| DBIOðð L ð1 2ð .ððð ð ð .ððð ð ð ð ð ð ð ð
| .
| .
| .
| JO-EVALUAT P ð2 ðð .ð35 ð ð .ððð 3 382 ð ð ð ð ð
| JO-TUNING- P ð2 ðð .ððð ð ð .ððð ð ð ð ð ð ð ð
| .
| .
| .
| QINTER QSYS ð23ð38 M ð2 ðð .ð13 ð ð .ððð 146 19 ð ð 32 ð 31
| QIWVPPJT QUSER ð2314ð B ð2 2ð .ððð ð ð .ððð 5 ð ð ð 4 ð ð
| QJOBSCD QSYS ð22993 S ð2 ðð .ððð ð ð .ððð ð ð ð ð ð ð ð
| QJVACMDSRV COOK ð23416 B ð4 26 4.113 ð ð .ððð 575 51 ð ð 1 ð 117
| QJVACMDSRV COOK ð23417 B ð4 26 4.14ð ð ð .ððð 565 51 ð ð 1 ð 117
| QJVACMDSRV GLMS ð23264 B ð4 26 15.496 ð ð .ððð 14ð 1ð ð ð 6 ð 44
| QLUR QSYS ð22995 S ð2 ðð .ððð ð ð .ððð ð ð ð ð ð ð ð
| QLUS QSYS ð22982 S ð2 ðð .ððð ð ð .ððð ð ð ð ð ð ð ð
| QLZPSERV QUSER ð23ð53 C ð2 2ð .ððð ð ð .ððð 6 ð ð ð 1 ð ð
| QMSF QMSF ð23ð78 B ð2 35 .ððð ð ð .ððð 6 ð ð ð 5 ð ð
| QNETWARE QSYS ð23ð77 B ð2 5ð .ðð2 ð ð .ððð 4 ð ð ð 2 ð ð
| QNFTP QSNADS ð2311ð B ð2 4ð .ððð ð ð .ððð 5 ð ð ð 4 ð ð
| QNMAPINGD QUSER ð23ð47 B ð2 25 .ððð ð ð .ððð 4 ð ð ð 1 ð ð
| QNMAREXECD QUSER ð23ð49 B ð2 25 .ððð ð ð .ððð 3 ð ð ð 1 ð ð
| QNNDIRQS QNOTES ð23ð55 B ð2 35 .ðð1 ð ð .ððð 3 ð ð ð 2 ð ð
| .
| .
| .
| QPADEVðð16 COOK ð23399 P ð4 2ð .ð52 19 69 .1ð5 118 2ð 14 ð ð ð 12
| QPADEVðð16 COOK ð23414 P ð4 2ð .473 9 133 .111 57 9 ð ð ð ð 21
| QPADEVðð17 COOK ð23398 P ð4 2ð .ð61 35 127 5.828 134 17 1ð ð ð ð 25
| QPADEVðð17 COOK ð23415 P ð4 2ð .481 9 137 .111 43 7 ð ð 1 ð 16
| QPADEVðð19 GLMS ð23262 P ð4 2ð .ð8ð 38 11ð .ð78 38 4 ð ð 2 ð 5
| QPADEVðð2ð GLMS ð23291 P ð4 2ð .ðð8 ð ð .ððð 18 3 ð ð 4 ð 5
| QPASVRP QSYS ð23ð26 B ð2 5ð .ððð ð ð .ððð 5 ð ð ð 2 ð ð
| QPASVRS QSYS ð23ð66 B ð2 3ð .ððð ð ð .ððð 2 ð ð ð 2 ð ð
| .
| .
| .
| Column Total Average
| --------------------------- ---------------- ----------------
| CPU Util 98.787
| Tns 115
| Tns /Hour 333
| Resp 1.834
| Sync Disk I/O 13,5ð8
| Async Disk I/O 39,4ð7
| Logical Disk I/O 728,485
| Cmn I/O ð
| PAG Fault 115
| Arith Ovrflw 4ð7,888
| Perm Write 25,444

Figure 7-16. Job Workload Activity

7-30 Performance Tools V4R2

 Storage Pool Activity–Sample
| Component Report 11/11/98 8:2ð:28
| Storage Pool Activity Page 15
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . : 11/11/98 ð7:58:ð1
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : 11/11/98 ð8:18:43
| Pool identifier . . . : ð5 Expert Cache . . . : 3
| Pool Avg ---------- Avg Per Second ----------- ----- Avg Per Minute -----
| Itv Size Act Total Rsp CPU ------ DB ------ ----- Non-DB ----- Act- Wait- Act-
| End (KB) Level Tns Time Util Faults Pages Faults Pages Wait Inel Inel
| ----- ----------- ----- ----------- ----- ---- ------- ------ ------- ------ ------ ------ ------
| ð8:ð3 66,ððð 3 ð .ðð 5.2 .1 2 .ð ð ð ð ð
| ð8:ð8 66,ððð 3 ð .ðð 5ð.8 .4 11 .ð ð ð ð ð
| ð8:13 66,ððð 3 ð .ðð 95.8 .ð 23 .ð ð ð ð ð
| ð8:18 66,ððð 3 ð .ðð 93.4 .ð 23 .ð ð ð ð ð
| ð8:18 66,ððð 3 ð .ðð 95.9 .ð 34 .ð ð ð ð ð
| .
| .
| .
| Itv End -- Interval end time (hour and minute)
| Pool Size (KB) -- Initial pool size in kilobytes (1ð24)
| Act Level -- Initial pool activity level
| Total Tns -- Number of transactions processed in this pool
| Avg Resp Time -- Average transaction response time
| CPU Util -- Percentage of available CPU time used by the job. This is the average of all processors
| DB Faults -- Database faults per second
| DB Pages -- Database pages per second
| Non-DB Faults -- Nondatabase faults per second
| Non-DB Pages -- Nondatabase pages per second
| Act-Wait -- Number of active-to-wait transitions per minute
| Wait-Inel -- Number of wait-to-ineligible transitions per minute
| Act-Inel -- Number of active-to-ineligible transitions per minute

Figure 7-17. Storage Pool Activity

Disk Activity–Sample

Chapter 7. Performance Reports—Manager Feature 7-31

| Component Report 11/11/98 8:2ð:28
| Disk Activity Page 16
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . : 11/11/98 ð7:58:ð1
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : 11/11/98 ð8:18:43
| ---------------- Average Disk Activity Per Hour ---------------- ----- Cache hit Statistics ----
| Srv ------------------ Disk Arm Seek Distance ---------------------- Device Controler Write -Disk Capacity-
| Unit Util Time ð 1/12 1/6 1/3 2/3 >2/3 Read Read Efficiency MB Percent
| ---- ---- ----- --------- --------- --------- --------- --------- --------- ------- --------- ---------- ------- -------
| ððð1 A 1.8 .ðð12 2,367 1,435 4ð6 75 574 95 .ð .ð .ð 446 22.6
| ððð1 B 1.5 .ððð9 2,375 1,897 5ð7 95 243 2 .ð .ð .ð 446 22.6
| ððð3 .4 .ððð2 3,156 87ð 435 1,131 1,618 217 17.ð .1 33.5 374 25.3
| ððð4 .9 .ððð3 3,457 1,395 69ð 983 1,757 4ð3 3ð.7 .ð 15.9 378 25.6
| ððð5 .8 .ððð4 3,272 887 51ð 936 1,746 362 18.6 .ð 14.9 374 25.3
| ððð6 .8 .ððð3 2,897 1,ð99 365 1,ð84 1,435 229 4ð.2 .1 39.4 381 25.8
| ððð7 2.ð .ððð2 6,329 2,294 1,ðð6 1,45ð 2,66ð 835 57.1 .1 59.4 814 25.8
| ððð8 .6 .ððð2 4,911 1,372 1,ð53 1,316 1,734 481 2ð.5 .1 49.8 8ð3 25.5
| ððð9 5.ð .ððð2 7,484 4,65ð 1,357 2,149 2,233 5ð1 73.9 .ð 81.8 814 25.8
| ðð1ð .5 .ððð2 4,832 2,ð39 76ð 1,433 1,644 3ð4 29.ð .1 2.2 811 25.7
| Column Total Average
| --------------------------- ---------------- ----------------
| Util 1.4
| Srv Time .ððð3
| Disk Arm Seek Distance
| ð 41,ð85
| 1/12 17,941
| 1/6 7,ð92
| 1/3 1ð,657
| 2/3 15,65ð
| >2/3 3,434
| Cache hit Statistics
| Device Read 52.8
| Controller Read .ð
| Write Efficiency 59.5
| Disk Capacity
| MB 5,641
| Percent 25.1
| Unit -- Disk arm identifier
| Util -- Drive utilization
| Srv Time -- Average service time per request in seconds
| Disk Arm Seek Distance -- Average seek distance distributions per hour
| ð -- Number of zero seeks
| 1/12 -- Number of seeks between ð and 1/12 of the disk
| 1/6 -- Number of seeks between 1/12 and 1/6 of the disk
| 1/3 -- Number of seeks between 1/6 and 1/3 of the disk
| 2/3 -- Number of seeks between 1/3 and 2/3 of the disk
| >2/3 -- Number of seeks greater than 2/ 3 of the disk
| Cache hit Stastistics --
| Device Read -- Percent of Device Read Hits for each arm
| Controller Read -- Percent of Controller Cache Read Hits for each arm
| Write Efficiency -- Percent of Efficiency of write Cache
| Disk Capacity -- Average amount of disk space used or available
| MB -- Millions of bytes available on the disk
| Percent -- Percent of space available on the disk

Figure 7-18. Disk Activity

IOP Utilizations–Sample

7-32 Performance Tools V4R2

| Component Report 9/15/98 1ð:51:38
| IOP Utilizations Page 29
| Sample Component Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| --- IOP Processor Util --- DASD -- KBytes Transmitted -- Available
| IOP Total Comm LWSC DASD Ops/Sec IOP System Storage Util 2
| -------------------- ----- ----- ----- ----- ----------- ----------- ----------- ----------- ------
| CCð1 (2623) .2 .ð 94 ð 1,772,ð96 .ð
| CCð2 (2623) .2 .ð 94 ð 1,772,ð96 .ð
| CCð3 (2626) 28.4 28.4 488,619 316,ð44 287,972 .ð
| CCð4 (2626) .1 .ð 94 ð 1,772,216 .ð
| CCð5 (2619) 3.5 3.5 .ð .ð 191,623 1,191 1,864,336 .ð
| CMBð1 (918B) 3.6 .ð .ð 3.5 1,5ð6 7,89ð,496 2,122,56ð .ð
| SIð2 (6533) 2.7 .ð .ð 1.7 3,645 29,324 6,485,936 .ð
| WSð1 (2661) .6 63 39
| IOP -- Resource name and model number for each communications, DASD,
| multifunction, and local work station IOP
| IOP Processor Util Total -- Total utilization for IOP
| IOP Processor Util Comm -- Utilization of IOP due to communications activity
| IOP Processor Util LWSC -- Utilization of IOP due to local work station activity
| IOP Processor Util DASD -- Utilization of IOP due to DASD activity
| DASD Ops/Sec -- Disk operations per second
| KBytes Transmitted IOP -- Total Kbytes transmitted from an IOP to the system across the bus
| KBytes Transmitted System -- Total Kbytes transmitted to the IOP from the system across the bus
| Available Storage -- The average number of bytes of free local storage in the IOP
| Util 2 -- Utilization of co-processor

Figure 7-19. IOP Utilizations

Local Work Stations–Response Time Buckets–Sample

| Component Report 9/15/98 1ð:51:38
| Local Work Stations - Response Time Buckets Page 3ð
| Sample Component Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| Ctl/Device Util IOP Name
| --------------- ---- ------------
| CTLð1 .6 WSð1
| ð- 1.ð 1.ð- 2.ð 2.ð- 4.ð 4.ð- 8.ð > 8.ð Rsp Time
| ------ --------- --------- --------- ------ --------
| DSPð1 ð ð 1 ð ð 3.ðð
| DSPð2 23 1 2 1 2 2.89
| Total Responses 23 1 3 1 2 2.9ð
| Ctl -- Controller identifier
| Device -- Device identifier
| Util -- Controller utilization
| IOP Name -- Input/Output processor resource name
| ð- 1.ð -- Number of response times in this range
| 1.ð- 2.ð -- Number of response times in this range
| 2.ð- 4.ð -- Number of response times in this range
| 4.ð- 8.ð -- Number of response times in this range
| > 8.ð -- Number of response times in this range
| Rsp time -- Average external response time (in seconds)
| for this workstation(s)

Figure 7-20. Local Work Stations – Response Time Buckets

Remote Work Stations–Response Time Buckets–Sample

Chapter 7. Performance Reports—Manager Feature 7-33

 Component Report 9/24/98 7:38:ð5
Remote Work Stations - Response Time Buckets Page 18
Sample Component Report
Member . . . : TEST2ð Model/Serial . : 5ðð-2142/1ð-317CD Main storage . . : 128.ð M Started . . . . : ð9/19/98 16:47:34
Library . . : RWSDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/19/98 17:12:36
Ctl/Device IOP Name
--------------- ------------
ABSYSTEM CCð2
ð- 1.ð 1.ð- 2.ð 2.ð- 4.ð 4.ð- 8.ð > 8.ð Rsp Time
------ --------- --------- --------- ------ --------
RCH5DSPð7 845 ð ð ð ð .ð2
Total Responses 845 ð ð ð ð .ð2
Ctl -- Controller identifier
Device -- Device identifier
IOP Name -- Input/Output processor resource name
ð- 1.ð -- Number of response times in this range
1.ð- 2.ð -- Number of response times in this range
2.ð- 4.ð -- Number of response times in this range
4.ð- 8.ð -- Number of response times in this range
> 8.ð -- Number of response times in this range
Rsp time -- Average external response time (in seconds)
for this workstation(s)

Figure 7-21. Remote Work Stations – Response Time Buckets

Exception Occurrence Summary and Interval Counts–Sample

| Component Report 9/15/98 1ð:51:38
| Exception Occurrence Summary and Interval Counts Page 31
| Sample Component Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| Exception Counts
| Exception
| Type Description Total
| ---------------- -------------------------------- -----------
| Size Size ð
| Binary Overflow Binary overflow ð
| Decimal Overflow Decimal overflow ð
| Flp Overflow Floating point overflow ð
| Decimal Data Decimal data ð
| Aut Lookup Authority lookup 4,811
| PAG Fault Process Access Group fault 3,819
| Seize Conflict Seize conflict 2,236
| Lock Conflict Lock conflict 344
| Verify Verify ð
| -------------------------------------------- Exceptions Per Second ---------------------------------------------
| Itv Binary Decimal Flp Decimal Aut PAG Seize Lock
| End Size Overflow Overflow Overflow Data Lookup Fault Conflict Conflict Verify
| ----- ------- -------- -------- -------- -------- -------- -------- -------- -------- --------
| ðð:31 .ð .ð .ð .ð .ð .ð .ð .ð .ð .ð
| ðð:46 .ð .ð .ð .ð .ð .1 .ð .ð .ð .ð
| ð1:ð1 .ð .ð .ð .ð .ð .ð .ð .ð .ð .ð
| .
| .
| .

Figure 7-22. Exception Occurrence Summary and Interval Counts

Database Journaling Summary–Sample

7-34 Performance Tools V4R2

| Component Report 9/15/98 1ð:51:38
| Data Base Journaling Summary Page 34
| Sample Component Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð9/1ð/98 23:5ð:ð8
| ------- Journal Operations ------ --- Journal Deposits --- Bundle Bundle --Exposed AP -- --Est Exposr --
| Itv User User System System User System System Writes Writes System Not Curr AP Not SMAPP
| End Starts Stops Starts Stops Total Total ToUser User System Jrnld Jrnld System Jrnld ReTune
| ----- ------ ------ ------ ------ ------- ------- ------- ------- -------- ------ ------ ------ ------ ------
| ðð:31 ð ð ð ð ð ð ð ð ð ð 25 19 19 ð
| ðð:46 ð ð ð ð 4 ð ð 4 ð ð 25 19 19 ð
| ð1:ð1 ð ð ð ð 34 ð ð 34 ð ð 25 19 19 ð
| ð1:16 ð ð ð ð ð ð ð ð ð ð 25 19 19 ð
| ð1:31 ð ð ð ð ð ð ð ð ð ð 25 19 19 ð
| .
| .
| .
| Itv End -- Interval end time (hour and minute)
| User Starts -- Start journal operations initiated by user
| User Stops -- Stop journal operations initiated by user
| System Starts -- Start journal operations initiated by system
| System Stops -- Stop journal operations initiated by system
| User Total -- Journal deposits resulting from user journaled
| objects
| System Total -- Journal deposits resulting from system journaled
| objects (total)
| System ToUser -- Journal deposits resulting from system journaled
| objects to user created journals
| Bundle Writes User -- Bundle writes to user created journals
| Bundle Writes System -- Bundle writes to internal system journals
| Exposed AP System Jrnld -- Exposed access paths currently being journaled
| by the system
| Exposed AP Not Jrnld -- Exposed access paths currently not being journaled
| Est Exposr Curr System -- System estimated access path recovery time
| exposure in minutes
| Est Exposr AP Not Jrnld -- System estimated access path recovery time
| exposure in minutes if no access
| paths were being journaled by the system
| SMAPP ReTune -- System Managed Access Path Protection tuning
| adjustments

Figure 7-23. Database Journaling Summary

Component Report Selection Criteria: Select Parameters–Sample

Component Report 11/2ð/95 1ð:ð6:31
Report Selection Criteria Page ðð14
Sample Component Report

Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð MB Started . . . . : ð5/ð3/96 14:59:32
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 3/ 7.ð Stopped . . . . : ð5/ð3/96 15:ð4:31

Select Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Select parameters were chosen.

Figure 7-24. Report Selection Criteria

Chapter 7. Performance Reports—Manager Feature 7-35

Component Report Selection Criteria: Omit Parameters–Sample
Omit Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Omit parameters were chosen.

Figure 7-25. Report Selection Criteria

Transaction Report

Printing the Transaction Report

Use the PRTTNSRPT command, or select option 3 (Transaction report) on the Print
Performance Reports display. When you use the PRTTNSRPT command, you can
choose to print three types of reports using the report type (RPTTYPE) parameter
Ÿ Job Summary Report (*SUMMARY)
Ÿ Transaction Report (*TNSACT)
Ÿ Transition Report (*TRSIT)

The Transaction and Transition Reports provide detailed information. So, when you
print these reports, use the selection values available on the PRTTNSRPT
command to select specific jobs, users, or time intervals. That way you can limit the
output to relevant information only.

The PRTTNSRPT command uses trace output from the STRPFRMON command.
The STRPFRMON command must be run first.

Notes:
1. In some instances, when a value is too large to fit in the allotted space, a 9 is
printed in each numeric field in the report. To see the actual value, you should
specify RPTTYPE(*FILE) on the PRTTNSRPT command.
2. The PRTTNSRPT command takes some CPU model values from the BEST/1
hardware table to do some calculations.

7-36 Performance Tools V4R2

 Collecting Trace Data
To create trace data:
1. Start the performance monitor (use the STRPFRMON command) and specify
tracing (TRACE(*ALL)).
2. When the monitor job completes, print the report. The value for the MBR
parameter on the print command should be the same as the value for the LIB
parameter on the print command.
Note: You can end the performance monitor before the specified interval has
ended by using the End Performance Monitor (ENDPFRMON)
command.

PRTTNSRPT Printer Files

The PRTTNSRPT command uses the following printer files:
File Description
QPSPDJS Job summary report output
QPSPDTS Transaction report output
QPSPDTD Transition report output

What Is the Transaction Report?

The following are the types of transaction reports:
Ÿ Job Summary Report provides general job information. Always request this
report first.
Ÿ Transaction Report provides detailed information about each transaction that
occurred in the job:
– Transaction response time
– Name of the program that is active at the time the transaction starts
– Processing unit time use
– Number of I/O requests
Ÿ Transition Report provides information similar to that of the Transaction
Report, but the data (for example, processing unit time, I/O requests) is shown
for each job state transition, rather than just the transitions shown when the job
is waiting for work station input. The detail shown in this report helps you to
determine the program that ran during a transition, or to determine when an
unsatisfied lock request occurred.

For samples of each section of the Transaction Report, see “Sample Transaction
Reports” on page 7-49.

For definitions of specific columns in the reports, see “Performance Report

Columns” on page 7-102.

Chapter 7. Performance Reports—Manager Feature 7-37

Job Summary Report
The Job Summary Report (RPTTYPE(*SUMMARY)) provides the following
sections:
Ÿ Job Summary
Ÿ System Summary Data
Ÿ Distribution of Transactions by CPU/Transaction
Ÿ Transaction Significance
Ÿ Interactive Transactions by 5-Minute Intervals
Ÿ Interactive Throughput by 5-Minute Intervals
Ÿ Interactive CPU Utilization by 5-Minute Intervals
Ÿ Interactive Response Time by 5-Minute Intervals
Ÿ Scatter Diagram of Interactive Transactions by 5-Minute Intervals
Ÿ Interactive Program Transaction Statistics
Ÿ Summary of Seize/Lock Conflicts by Object
Ÿ Report Selection Criteria

Job Summary Section

The Job Summary section of the Job Summary Report shows the following informa-
tion for each job in the system:
Ÿ The name and type of job (for example, interactive, batch)
Ÿ The number of transactions in the job
Ÿ The average transaction response time
Ÿ The average processing unit time per transaction
Ÿ The average number of disk I/O requests per transaction
Ÿ The number of lock waits
Ÿ The number of seize conflicts
Ÿ The key/think time per transaction

If the Job Summary section shows jobs that have high response times, high disk
I/O activity, high processing unit utilization, or a number of lock requests, use the
Transaction Report to investigate further.

If the number of seizes or number of conflicts (Number Sze Cft or Number Lck
Cft columns on this report) is “high,” look at the Transaction or Transition reports
for the job to see how long the conflict lasted, the job that held the object, the name
and type of object being held, and what the job was waiting for.

The exact meaning of the term “high” is dependent on the application. One
example is the number of lock-waits. An application that has many users
accessing a database at the same time could, under normal conditions, have
numerous lock-waits.

You must evaluate each situation individually. If the values are difficult to explain
(an application should have very few locks and yet many are reported), then further

7-38 Performance Tools V4R2

analysis will be required. The Transaction and Transition Reports can help in this
analysis.

System Summary Data Section (First Part)

The first part of the System Summary Data section of the Job Summary Report
includes the following:

Ÿ Trace Periods for Trace Date

Ÿ CPU by Priority for All Jobs for Total Trace Period

See the sample report shown in Figure 7-27 on page 7-51.

System Summary Data (Second Part)

The second part of the System Summary Data section of the Job Summary Report
includes the following:
Ÿ CPU and Disk I/O per Job Type for All Jobs for Total Trace Period
Ÿ Interactive Transaction Averages by Job Type

See the sample report shown in Figure 7-28 on page 7-51.

System Summary Data (Third Part)

The Analysis by Interactive Transaction Categories part of the System Summary
Data section provides a breakdown of the transactions into the categories very
simple, simple, medium, and complex, relative to their processing unit utilization.

The boundary values that are used to categorize the transactions by processing
unit model were updated to more accurately reflect a typical customer workload.
The boundary values have almost doubled. For the typical customer workload, this
update causes the number of transactions categorized as simple and medium to
increase, and those categorized as complex and very complex to decrease. This
does not change the data itself or how it is collected. The update only changes how
individual transactions are categorized by the Transaction Report.
Note: The Total/Avg is only a total or average of the simple, medium, and
complex categories. The very simple category is a part of the simple cate-
gory. The very complex category is a part of the complex category.

These transaction categories depend on the processing unit model. They are intro-
duced here and in some of the following reports as a way to highlight the differ-
ences that exist in the work being done on the system.

When you are considering adding new applications, determine the new application’s
transaction characteristics. For example, determine if a high volume of complex
transactions is typical with this new application. By analyzing the transaction char-
acteristics of new applications, you may be able to foresee the need to acquire
additional hardware resources for the new application.

If you obtain a new application from a supplier, it is reasonable to ask for informa-
tion about the application’s transaction characteristics.

The Analysis by Interactive Response Time part of the System Summary Data
section provides transaction information sorted by response time categories.

Chapter 7. Performance Reports—Manager Feature 7-39

 The Analysis by Interactive Key/Think Time part of the System Summary Data
section provides information about the key/think time.

See the sample report shown in Figure 7-29 on page 7-52.

Distribution by CPU/Transaction Section

The Distribution of Transactions by CPU/Transaction section of the Job Summary
Report provides a graphical view of the distribution of simple, medium, and complex
transactions. This chart shows the number of transactions versus the processing
unit time per transaction in seconds.

See the sample report shown in Figure 7-30 on page 7-53.

Transaction Significance Section

The Transaction Significance section of the Job Summary Report provides a graph-
ical view of the processing unit use, categorized by simple, medium, and complex
transactions. This chart shows the percent of available processing unit time used
versus the processing unit time per transaction in seconds.

See the sample report shown in Figure 7-31 on page 7-54.

Transactions by Intervals Section

The Interactive Transactions by 5-Minute Intervals section of the Job Summary
Report provides a count of the number of active jobs during a 5-minute interval that
performed at least one transaction. It also shows the number of jobs that were
signed on and off during the 5-minute intervals. Transaction rates per 5-minute
intervals are shown in several different formats.

See the sample report shown in Figure 7-32 on page 7-54.

Interactive Throughput Section

The Interactive Throughput by 5-Minute Intervals section of the Job Summary
Report gives simple, medium, and complex transactions relative to the number of
transactions according to an interval end time.

See the sample report shown in Figure 7-33 on page 7-55.

Interactive CPU Utilization Section

The Interactive CPU Utilization by 5-Minute Intervals section of the Job Summary
Report gives simple, medium, and complex transactions relative to their processing
unit utilization.

See the sample report shown in Figure 7-34 on page 7-55.

Interactive Response Time Section

The Interactive Response Time by 5-Minute Intervals section of the Job Summary
Report gives the response components relative to the resulting response time.

See the sample report shown in Figure 7-35 on page 7-55.

7-40 Performance Tools V4R2

 Scatter Diagram Section
The Scatter Diagram section of the Job Summary Report gives the average of
measured response times for 5-minute intervals compared to transaction rates.

See the sample report shown in Figure 7-36 on page 7-56.

Interactive Program Transaction Statistics Section

The Interactive Program Transaction Statistics section of the Job Summary Report
arranges the programs by the number of transactions associated with the pro-
grams.

See the sample report shown in Figure 7-37 on page 7-57.

Seize/Lock Conflicts by Object Section

The Summary of Seize/Lock Conflicts by Object section of the Job Summary
Report gives information about the locks and seizes associated with objects. The
unnamed object, shown as ADDR 00000E00, is the Licensed Internal Code data-
base in-use table. It often appears in this report when there are a high number of
database file opens and closes.

See the sample report shown in Figure 7-38 on page 7-57.

Special System Information

In general, the information identifies exceptional conditions and events that occur
over the measurement period. If you analyze these exceptions, you might find jobs
and programs you need to examine. A summary of these sections of the Job
Summary Report follows.
Ÿ Priority-Jobtype-Pool Statistics section
Ÿ Job Statistics section
Ÿ Interactive Program Statistics section
Ÿ Individual Transaction Statistics section
Ÿ Longest Seize/Lock Conflicts section
Ÿ Longest Holders of Seize/Lock Conflicts section
Ÿ Batch Job Analysis section
| Ÿ Concurrent Batch Job Statistics section
Ÿ Report Selection Criteria section

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

Chapter 7. Performance Reports—Manager Feature 7-41

Priority-Jobtype-Pool Statistics Section
The Priority-Jobtype-Pool Statistics section of the Job Summary Report shows the
total processing unit seconds and physical I/O requests for each category of
priority-jobtype and pool combination recorded during the overall test period. The
number of total transactions is shown for job type I only.

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-39 on page 7-58.

Job Statistics Section

The Job Statistics section of the Job Summary Report shows the 10 jobs with the:
Ÿ Most transactions (shown in Figure 7-40 on page 7-58)
Ÿ Largest average response time
Ÿ Largest average processing unit time per transaction
Ÿ Largest synchronous disk I/O per transaction
A synchronous disk I/O is a disk access operation that must complete before
program operation can continue.
Ÿ Largest asynchronous disk I/O per transaction.
An asynchronous disk I/O is a disk access operation that is not expected to
complete before program operation can continue.
Ÿ Most seize conflicts
Ÿ Most record lock conflicts
Ÿ Most active-to-ineligible occurrences
Ÿ Most wait-to-ineligible occurrences
Ÿ Most event wait occurrences

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-40 on page 7-58.

Interactive Program Statistics Section

The Interactive Program Statistics section of the Job Summary Report gives addi-
tional program information showing the top 10 programs with the largest average:
Ÿ Processing unit time per transaction (shown in Figure 7-41 on page 7-59)
Ÿ Synchronous disk I/O per transaction
Ÿ Asynchronous disk I/O per transaction
Ÿ Response time per transaction
Ÿ Synchronous database reads per transaction
Ÿ Synchronous database writes per transaction
Ÿ Synchronous nondatabase reads per transaction

7-42 Performance Tools V4R2

 Ÿ Synchronous nondatabase writes per transaction

See the sample report shown in Figure 7-41 on page 7-59.

Individual Transaction Statistics Section

The Individual Transaction Statistics section of the Job Summary Report lists the 10
transactions with the least or most:
Ÿ Response time (shown in Figure 7-42 on page 7-59)
Ÿ Processing unit service time
Ÿ Total synchronous disk I/O
Ÿ Total asynchronous disk I/O
Ÿ Synchronous database reads
Ÿ Synchronous database writes
Ÿ Synchronous nondatabase reads
Ÿ Synchronous nondatabase writes
Ÿ Asynchronous database reads
Ÿ Asynchronous database writes
Ÿ Asynchronous nondatabase reads
Ÿ Asynchronous nondatabase writes
Ÿ Short-wait-extended time
Ÿ Short-wait time
Ÿ Lock-wait time
Ÿ Excessive activity-level wait time
Ÿ Active time
Ÿ Binary overflow exceptions
Ÿ Decimal overflow exceptions
Ÿ Floating point overflow exceptions
Ÿ Process access group fault exceptions
Ÿ Permanent writes

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-42 on page 7-59

Longest Seize/Lock Conflicts Section

The Longest Seize/Lock Conflicts section of the Job Summary Report shows the 30
longest lock or seize conflicts during the trace period.

Chapter 7. Performance Reports—Manager Feature 7-43

 To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-43 on page 7-60.

Longest Holders of Seize/Lock Conflicts Section

The Longest Holders of Seize/Lock Conflicts section of the Job Summary Report
shows the holders of the longest lock or seize conflicts for all job types during the
trace period.

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-44 on page 7-60.

Batch Job Analysis Section

Note: The Batch Job Analysis section does not print if you also specify a value on
the select job (SLTJOB) parameter or the omit job (OMTJOB) parameter.

The Batch Job Analysis section of the Job Summary Report shows information on
the batch job workload during the trace period.

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-45 on page 7-60.

Concurrent Batch Job Statistics

The Concurrent Batch Job Statistics section of the Job Summary Report shows
information on the batch job workload during the trace period according to job sets.

By looking at the first lines for a particular priority, you can quickly determine if the
system was fully utilizing all available batch activity levels during the trace period.

The performance monitor begins identifying concurrent jobs when the monitor is
started. All jobs that are currently active are assigned to a job set. There will often
be several jobs that are continuously active during the trace period, such as
autostart jobs for SNADS and the QPFRMON job itself.

If another job starts during the trace period and none of the original jobs have
ended, it is assigned to a new job set. If a job ends and another job of the same
priority starts, the new job is considered to be a second job in the same job set.

For example, if the job queue entry for QBATCH has a MAXACT parameter of 3
and you submit 8 jobs to QBATCH during the trace period, there will probably be 3
job sets on the report with a total of 8 jobs shared between them.

The threads are sorted by job priority. Thus, for the above example where the first
thread was running for a total of 8 minutes and 50 seconds and the second thread
was running for a total of 6 minutes and 55 seconds, the order of reporting shows
the statistics for the second thread, then the third, and then the first and assigns
them sequential numbers.

7-44 Performance Tools V4R2

 To Print
Use OPTION(*SS) on the PRTTNSRPT command.
Note: The Concurrent Batch Job Statistics section does not print if you also
specify a value on the select job (SLTJOB) parameter or the omit job
(OMTJOB) parameter.

See the sample report shown in Figure 7-46 on page 7-61.

Report Selection Criteria Section

The Report Selection Criteria section of the Job Summary Report gives the
selection values you chose to produce the report.

Use the SELECT parameters on the Report Selection Criteria Report to select
pools, jobs, user IDs or functional areas. Or use the OMIT parameters to omit them

If you did not use SELECT parameters, the No Select parameters were chosen
message appears.

If you did not use OMIT parameters, the No Omit parameters were chosen
message appears.

The options which were selected are also given.

To Print
Use OPTION(*SS) on the PRTTNSRPT command.

See the sample report shown in Figure 7-47 on page 7-61.

Transaction Report
The Transaction Report (RPTTYPE(*TNSACT)) provides detailed information about
each transaction that occurred in the job:
Ÿ Transaction response time
Ÿ Name of the program that is active at the time the transaction starts
Ÿ Processing unit time use
Ÿ Number of I/O requests

The Transaction Report output has two parts:

Ÿ The details, which show data about each transaction in the job
Ÿ The summary, which shows data about overall job operation

To Print
Use RPTTYPE(*TNSACT) on the PRTTNSRPT command.

See the sample report shown in Figure 7-48 on page 7-62.

Chapter 7. Performance Reports—Manager Feature 7-45

Job Summary Data
The Job Summary Data section of the Transaction Report includes averages of the
job data. Some of this information is also found in the Job Summary section of the
Job Summary Report.

Transition Report
The Transition Report (RPTTYPE(*TRSIT)) provides information similar to that of
the Transaction Report, but the data (for example, processing unit time, I/O
requests) is shown for each job state transition, rather than just the transitions
shown when the job is waiting for work station input. The detail shown in this report
helps you to determine the program that ran during a transition, or to determine
when an unsatisfied lock request occurred.

The Transition Report is composed of two sections:

Ÿ Transition Detail, which shows each state transition made by the job (going
from one state to another, such as active-to-ineligible)
Ÿ Summary, which shows the same data as the summary output from the Trans-
action Report

To Print
Use RPTTYPE(*TRSIT) on the PRTTNSRPT command.

See the sample report shown in Figure 7-49 on page 7-62.

Transition Detail
The job transitions for jobs using data queues are in the State column of the Tran-
sition Detail report. If a job uses data queues (CALL QSNDDTAQ or CALL
QRCVDTAQ), each access to the queue is marked with an EOT2-SOT2 pair. If
data is received by a queue when the data queue currently has no entries, the
transition detail report shows a job state of wait (W in the STATE column), but leaves
the job in the activity level up to a short wait time (2 seconds) or until the interval
time set for the time slice end.

When either the time-out value of the QRCVDTAQ API ends or data is returned
from the queue, the transition report records an -->A in the STATE column.

If a job is doing interactive I/O operations to an ICF file, the transition detail records
a W<-- and -->A pair under the STATE column for start (W) and completion (A) of
the write or read operation. For example, if the job is doing APPC I/O operations
within an interactive transition with a display device:
Time stamp SOT1
Time stamp W<--
Time stamp -->A
Time stamp W<--
Time stamp -->A
| job processing
Time stamp EOR1
Time stamp EOT1

7-46 Performance Tools V4R2

If the wait code column has an EORn, EOTn, or SOTn, the two program names on
the left are filled in with information from the transaction boundary trace record, and
the two program names on the right are blank.

The program name under Last contains the following information:

Transaction Name
Display I/O Display device
Data queue Data queue library
MRT Display device
Pass-through Device description
The program name under Second contains the following information:
Transaction Name
Display I/O Display file
Data queue Data queue
MRT Display file
Source pass-through Target control point
Target pass-through Source control point
WSF target pass-through Controller description

The values ADR=000000 or ADR=UNKWN can also appear as the program name.
The ADR=000000 occurs when there was no program active at that level in the job
when the trace record was created. ADR=UNKWN indicates that the program did
not exist on the system at the time the trace record data was dumped to a data-
base file. This happens if you have deleted (or replaced) the program before
dumping the file. The program names are put into the trace record when the
monitor ends and the trace data is put into a database file or when the Dump Trace
(DMPTRC) command is used.

Summary
The summary section of the Transition Report shows the same information as the
summary section of the Transaction Report, described in “Job Summary Data” on
page 7-46.

Table 7-3 on page 7-48 shows jobs with a W← (wait) job state and 130 for a
decimal qualifier. The job went from an active-to-wait state and dropped from the
activity level (this defines the end of a transaction in the report).

Chapter 7. Performance Reports—Manager Feature 7-47

 Table 7-3. W← Job States and Decimal Qualifiers
State W A I Wait Code Decimal Qualifier Description
W← – 130 Dequeue wait (Flag X'64).
W← EVT 130 Wait on event (Flag X'A4).
W← LKW 130 Lock wait (Flag X'34').
W← HDW 130 Hold wait (Flag X'2C').
Note: For the W← entry (going to long wait and not holding an activity level position) the WRITES value includes
the I/O to write the PAG to disk as well as any other write operations that have occurred since the last
trace entry. You can verify this by comparing it to output from extended trace job by looking at the
WRITES values across a WAIT entry (on the TRCJOB command).
You cannot exactly compare the times in the MPL trace data records with the times from TRCJOB or storage
management trace. Each uses a different method to convert to an HH:MM:SS.SSS value from an 8-byte
hexadecimal clock.

Table 7-4 shows jobs with a W (wait) job state and a decimal qualifier of 134. The
job went from active-to-wait state but stayed in the activity level (for example, a
short wait).

Table 7-4. W Job States and Decimal Qualifiers

State W A I Wait Code Decimal Qualifier Description
W – 134 Dequeue wait (Flag X'64').
W EVT 134 Wait on event (Flag X'A4').

Table 7-5 shows jobs with an I (ineligible) job state.

Table 7-5. I Job States and Decimal Qualifiers

State W A I Wait Code Decimal Qualifier Description
→I – 128 A new task cannot start.
→I – 132 Wait-to-ineligible transition.
→I TSE 136 Active-to-ineligible (time slice end).

Table 7-6 on page 7-49 shows jobs with an A (active) job state.

7-48 Performance Tools V4R2

 Table 7-6. A Job States and Decimal Qualifiers
State W A I Wait Code Decimal Qualifier Description
A – 142 Wait-to-active but already in the activity
level.
A← – 129 Ineligible-to-active transition.
A – 131 Message received while the task was in
the current MPL.
A – 133 Dequeue after time-out, task in current
MPL when message received.
→A – 135 Job went from wait-to-active state (this
defines the beginning of a transaction in
the report).
A WTO 137 Wait timed out.
A TSE 139 Active-to-active (external time slice end
based on time slice value in class) For
example, a time slice end occurred and
no jobs were waiting for an activity level.
A TSE 145 Active-to-active (internal time slice end
based on time slice value defined on
STRPFRMON).

Sample Transaction Reports

See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Job Summary–Sample

Chapter 7. Performance Reports—Manager Feature 7-49

| Job Summary Report 11/11/98 8:35:42
| Job Summary Page ððð5
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . :11/11/98 ð7:58:1ð
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :11/11/98 ð8:18:45
| \On/Off\ T P P Tot Response Sec CPU Sec ---- Average DIO/Transaction ----- Number K/T
| Job User Name/ Job y t r Nbr ------------- ------------------ ------ Synchronous ----- --Async-- Cft /Tns
| Name Thread Number Pl p y g Tns Avg Max Util Avg Max DBR NDBR Wrt Sum Max Sum Max Lck Sze Sec
| ---------- ---------- ------ -- -- -- - ---- ------ ------ ---- ------ ------ ---- ---- ---- ---- ---- ---- ---- --- --- -----
| QPADEVðð1ð ITSCID38 \ð23413 ð4 I 2ð Y 6 .1 .3 .1 .ð7 .18 3 1 4 7 2 4 2
| QPADEVðð16 COOK \ð23414 ð4 I 2ð Y 44 1.ð 25.6 .4 .ð3 .54 1 1 22 3 3
| QPADEVðð17 COOK \ð23415 ð4 I 2ð Y 51 .5 7.5 .4 .ð2 .54 1 1 18 5 2
| QJVACMDSRV COOK ð23416 ð4 BD 2ð 7.3 1.9ð 574 51
| QJVACMDSRV COOK ð23417 ð4 BD 2ð 7.5 1.89 565 51
| QJVACMDSRV ððððððð6 ð23416 ð4 BD 2ð 54.9 .ð5 1
| QJVACMDSRV ððððððð7 ð23416 ð4 BD 2ð 1.6 .16 1
| QJVACMDSRV ððððððð8 ð23416 ð4 BD 2ð 1.ð .16
| QJVACMDSRV ððððððð9 ð23417 ð4 BD 2ð 52.ð .ð3
| QJVACMDSRV ðððððððA ð23417 ð4 BD 2ð 1.6 .16
| QJVACMDSRV ðððððððB ð23417 ð4 BD 2ð 1.ð .17 1
| .
| .
| .
| DOMINOð2 QSYS ð24859 ð2 M ðð
| QNNINSTS QNOTES ð2486ð ð2 B 2ð
| SERVER QNOTES ð24861 ð2 BD 2ð .ð3
| SERVER ðððððð12 ð24858 ð2 BD 2ð .13
| SERVER ðððððð13 ð24858 ð2 BD 2ð .1 .43
| REPLICA QNOTES ð24862 ð2 BD 2ð .ð5 1
| ROUTER QNOTES ð24863 ð2 BD 2ð .21 1
| UPDATE QNOTES ð24864 ð2 BD 2ð .ð6 1
| STATS QNOTES ð24865 ð2 BD 2ð .5 1.24 1
| .
| .
| .
| SERVER ðððððð12 ð24861 ð2 BD 2ð .13
| EVENT QNOTES ð24875 ð2 BD 2ð .ð8 1
| AMGR QNOTES ð24876 ð2 BD 2ð .21
| REPORT QNOTES ð24877 ð2 BD 2ð .15 13
| EVENT ðððððð3ð ð24875 ð2 BD 2ð .19 3
| EVENT ðððððð31 ð24875 ð2 BD 2ð .1 .4ð 1
| EVENT ðððððð32 ð24875 ð2 BD 2ð .1 .36
| EVENT ðððððð33 ð24875 ð2 BD 2ð .1 .37
| EVENT ðððððð34 ð24875 ð2 BD 2ð .1 .36
| EVENT ðððððð35 ð24875 ð2 BD 2ð .ð6 2
| EVENT ðððððð36 ð24875 ð2 BD 2ð .1 .36

Figure 7-26. Job Summary: Job Summary

System Summary Data (First Part)–Sample

7-50 Performance Tools V4R2

| Job Summary Report 9/ð5/98 14:ð8:5ð
| System Summary Data Page ðð8ð
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| TRACE PERIODS FOR TRACE DATE.
| Elapsed
| Started Stopped Seconds
| -------- -------- ---------
| 12.47.42 12.53.25 343
| CPU BY PRIORITY FOR ALL JOBS FOR TOTAL TRACE PERIOD.
| CPU Cum CPU CPU
| Pty CPU Util Util QM
| ---- -------------- ------ ------- -------
| ðð 65.354 19.ð5 19.ð5 1.235
| ð9 19.ð5 1.235
| 1ð 19.ð5 1.235
| 16 19.ð5 1.235
| 2ð 84.7ð6 24.69 43.74 1.777
| 25 .ð38 .ð1 43.75 1.777
| 3ð .ð31 43.75 1.777
| 35 .ð23 43.75 1.777
| 36 43.75 1.777
| 4ð .ð77 .ð2 43.77 1.778
| 49 43.77 1.778
| 5ð .125 .ð3 43.8ð 1.779
| 52 .ð29 43.8ð 1.779
| 6ð 43.8ð 1.779
| 68 2.32ð .67 44.47 1.8ðð
| 84 .396 .11 44.58 1.8ð4
| 98 .ðð9 44.58 1.8ð4

Figure 7-27. Job Summary: System Summary Data - 1

System Summary Data (Second Part)–Sample

| Job Summary Report 9/ð5/98 14:ð8:5ð
| System Summary Data Page ðð81
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
| CPU AND DISK I/O PER JOB TYPE FOR ALL JOBS FOR TOTAL TRACE PERIOD. \ USE CHGJOBTYP COMMAND FOR BLANK JOB TYPES \
| \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
| Job Nbr CPU CPU --Disk I/O Requests-- CPU Sec/ Sync I/O
| Type Jobs Seconds Util Sync Async Sync DIO /Elp Sec
| --------------- ------- ---------- ----- --------- --------- -------- --------
| INTERACTIVE 8 1.3 .4 474 1 .ðð27 1.4
| BATCH A,B,C,D,X 84 12.6 3.7 15ð7 378 .ðð84 4.4
| SPOOL WTR/RDR ð .ð .ð ð ð .ðððð .ð
| SYSTEM JOBS 3458 84.7 24.7 1459 27 .ð581 4.3
| SYSTEM TASKS 174 54.4 15.9 872 25593 .ð624 2.5
| --------------- -------- ------------ ----- --------- --------- -------- --------
| \\ TOTALS \\ 3724 153.ð 44.7 4312 25999 .ð355 12.6
| DATA FOR SELECTED TIME INTERVAL (OR TOTAL TRACE PERIOD IF NO TIME SELECTION).
| INTERACTIVE TRANSACTION AVERAGES BY JOB TYPE.
| T Avg CPU/ ----- Sync Disk I/O Rqs/Tns ----- Async W-I Excp Key/ Active Est
| y Nbr Nbr Pct Tns Rsp Tns DB DB NDB NDB DIO Wait Wait Think K/T Of
| p Prg Jobs Tns Tns /Hour (Sec) (Sec) Read Write Read Write Sum /Tns /Tns /Tns /Tns /Tns AWS
| -- --- ---- ------ ----- ----- ------ ------ ----- ----- ----- ----- ----- ----- ------ ------- ------ ------ ---
| I YES 8 46 1ðð.ð 482 .354 .ð29 ð ð 1ð ð 1ð ð .ððð .ðð1 6.ð64 6.ð64 1
| EXCEPTIONAL WAIT BREAKDOWN BY JOB TYPE.
| A-I Short Short Seize Lock Event Excs EM327ð DDM Svr Other
| Wait Wait WaitX Wait Wait Wait ACTM Wait Wait Wait
| Type Purge /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns
| ---- ----- ------ ------ ------ ------ ------ ------ ------ ------ ------- ------
| I YES .ððð .ððð .ððð .ððð .ððð .ððð .ðð1 .ððð .ððð .ððð

Figure 7-28. Job Summary: System Summary Data - 2

Chapter 7. Performance Reports—Manager Feature 7-51

 System Summary Data (Third Part)–Sample
| Job Summary Report 9/ð5/98 14:ð8:5ð
| System Summary Data Page ðð82
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| ANALYSIS BY INTERACTIVE TRANSACTION CATEGORIES.
| Avg Cum ----- Sync Disk I/O Rqs/Tns ----- Async Avg Excp Avg Est
| CPU CPU CPU DB DB NDB NDB DIO Nbr Pct Rsp Wait K/T Of
| Category /Tns Util Util Read Write Read Write Sum /Tns Tns Tns /Tns /Tns /Tns AWS
| ---------------- ------- ----- ----- ----- ----- ----- ----- ----- ----- ----- ----- ------- ------- ------- ---
| VERY SIMPLE VS .ð18 .2 39 84.8 .ð31 .ðð1 1.113 1
| ----------------
| \\ SIMPLE S .ð18 .2 .2 1 1 44 95.7 .ð49 .ðð1 1.255 1
| -Boundary- .ð79
| \\ MEDIUM M .2
| -Boundary- .1ð7
| \\ COMPLEX X .261 .2 .4 224 224 1 2 4.3 7.ð56 111.852 1
| ----------------
| VERY COMPLEX VX
| ---------------- ------- ----- ----- ----- ----- ----- ----- ----- ----- ----- ----- ------- ------- ------- ---
| Total/Avg of \\ .ð29 1ð 1ð 46 1ðð.ð .354 .ðð1 6.ð64 1
| ANALYSIS BY INTERACTIVE RESPONSE TIME.
| Avg Cum Avg Cum ----- Sync Disk I/O Rqs/Tns ----- Async Excp Avg
| Rsp Nbr Pct Pct CPU CPU CPU DB DB NDB NDB DIO Wait K/T
| Category /Tns Tns Tns Tns /Tns Util Util Read Write Read Write Sum /Tns /Tns /Tns
| ---------------- ------- ----- ----- ----- ------- ----- ----- ----- ----- ----- ----- ----- ----- ------- -------
| Sub-Second .ð49 44 95.7 95.7 .ð18 .2 .2 1 1 .ðð1 1.255
| 1 - 1.999 Sec 95.7 .2
| 2 - 2.999 Sec 95.7 .2
| 3 - 4.999 Sec 95.7 .2
| 5 - 9.999 Sec 7.ð56 2 4.3 1ðð.ð .261 .2 .4 224 224 1 111.852
| GE 1ð Seconds 1ðð.ð .4
| ANALYSIS BY INTERACTIVE KEY/THINK TIME.
| Avg Cum Avg Cum ----- Sync Disk I/O Rqs/Tns ----- Async Avg Excp
| K/T Nbr Pct Pct CPU CPU CPU DB DB NDB NDB DIO Rsp Wait
| Category /Tns Tns Tns Tns /Tns Util Util Read Write Read Write Sum /Tns /Tns /Tns
| ---------------- ------- ----- ----- ----- ------- ----- ----- ----- ----- ----- ----- ----- ----- ------- -------
| LT 2 Seconds .542 36 78.3 78.3 .ð18 .2 .2 .ð32 .ðð1
| 2 - 14.999 Sec 4.487 9 19.6 97.9 .ð54 .1 .3 3ð 3ð .957
| 15 - 29.999 Sec 97.9 .3
| 3ð - 59.999 Sec 97.9 .3
| 6ð - 299.999 Sec 219.ð4ð 1 2.2 1ðð.1 .2ð8 .1 .4 197 197 6.54ð
| GE 3ðð Seconds 1ðð.1 .4

Figure 7-29. Job Summary: System Summary Data - 3

Distribution of Simple, Medium, and Complex Processing Unit

Transactions–Sample

7-52 Performance Tools V4R2

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Distribution of Transactions by CPU/Transaction Page ðð83
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Number
| Of
| Tns
| ------
| 25+
| 24+
| 23+
| 22+
| 21+
| 2ð+
| 19+
| 18+
| 17+
| 16+
| 15+
| 14+
| 13+
| 12+ S
| 11+ S
| 1ð+ SS S
| 9+ SS S
| 8+ SS S
| 7+ SS S
| 6+ SS S
| 5+ SS S
| 4+ SSS S
| 3+ SSSSS S
| 2+ SSSSS S
| 1+ SSSSSSS X X
| 1- SSSSSSS X X
| +----+----+----+----+----+----+----+----+----+----+----+----+----+----+ +++
| .ððð .ð5ð .1ðð .15ð .2ðð .25ð .3ðð .35ð
| CPU/Transaction (Seconds)
| Transaction Categories:
| S = Simple Transactions
| m = Medium Transactions
| X = Complex Transactions

Figure 7-30. Job Summary: Distribution of Processing Unit Transactions

Transaction Significance–Sample

Chapter 7. Performance Reports—Manager Feature 7-53

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Transaction Significance Page ðð84
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Percent
| CPU
| -------
| .1+ X
| .1+ X
| .1+ X
| .1+ S X
| .1+ S X
| .1+ S X
| .1+ S X
| .1+ S X
| .1+ S S X X
| .1+ S S X X
| .1+ S S X X
| .ð+ S S X X
| .ð+ S S X X
| .ð+ S S X X
| .ð+ S S X X
| .ð+ S S X X
| .ð+ SS S S X X
| .ð+ SS S S X X
| .ð+ SS S S X X
| .ð+ SS S S X X
| .ð+ SSSS S X X
| .ð+ SSSS S X X
| .ð+ SSSS S X X
| .ð+ SSSSSS X X
| .ð+ SSSSSSS X X
| .ð- SSSSSSS X X
| +----+----+----+----+----+----+----+----+----+----+----+----+----+----+ +++
| .ððð .ð5ð .1ðð .15ð .2ðð .25ð .3ðð .35ð
| CPU/Transaction (Seconds)
| Transaction Categories:
| S = Simple Transactions
| m = Medium Transactions
| X = Complex Transactions

Figure 7-31. Job Summary: Transaction Significance

Interactive Transactions by 5-Minute Intervals–Sample

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Interactive Transactions by 5 Minute Intervals Page ðð85
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| --- Pct Of Tns --- Pct CPU By Nbr Nbr Sync Async Avg Excp Pct Seize Active Est
| Itv Active Nbr Tns --- Categories --- Categories Sign Sign DIO DIO Rsp Wait Ex-Wt Wait K/T Of
| End Jobs Tns /Hour %VS\ %S %M %X \%VX %S %M %X offs ons /Tns /Tns /Tns /Tns /Rsp /Tns /Tns AWS
| ----- ------ ----- ----- ---- --+--+-- ---- ---+--+--- ---- ---- ----- ----- ------- ------- ----- ------ ------ ---
| \\\
| 12.5ð\ 1 43 516 91\1ðð ð ð \ ð ð ð ð .ð35 .814
| 12.55\ 1 3 36 ð\ 33 ð 67 \ ð ð ð ð 153 4.918 75.ððð
| \ Denotes Partial Interval Data

Figure 7-32. Job Summary: Interactive Transactions by 5-Minute Intervals

Interactive Throughput by 5-Minute Intervals–Sample

7-54 Performance Tools V4R2

 Job Summary Report 5/ð7/98 13:52:1ð
Interactive Throughput by 5 Minute Intervals Page ðð1ð
Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 98 14:59:44
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5 ð3 98 15:ð4:36
Number Of Transactions Per Hour
Itv
End
ð 4ðð 8ðð 12ðð 16ðð 2ððð 24ðð 28ðð
----- +----+----+----+----+----+----+----+----+----+----+----+----+----+----+
\\\
15/ð5 XXXXXXXXXXXXXXXXXXSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
Throughput Components:
S = Simple Transactions
m = Medium Transactions
X = Complex Transactions

Figure 7-33. Job Summary: Interactive Throughput by 5-Minute Intervals

Interactive CPU Utilization by 5-Minute Intervals–Sample

Job Summary Report 5/ð7/98 13:52:1ð
Interactive CPU Utilization by 5 Minute Intervals Page ðð11
Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 98 14:59:44
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5 ð3 98 15:ð4:36
Percent CPU Utilization
Itv
End ð 1ð 2ð 3ð 4ð 5ð 6ð 7ð 8ð 9ð 1ðð
----- +----+----+----+----+----+----+----+----+----+----+
\\\
15/ð5 XXXX
CPU Components:
S = Simple Transactions
m = Medium Transactions
X = Complex Transactions

Figure 7-34. Job Summary: Interactive CPU Utilization by 5-Minute Intervals

Interactive Response Time by 5-Minute Intervals–Sample

Job Summary Report 5/ð7/968 13:52:1ð
Interactive Response Time by 5 Minute Intervals Page ðð12
Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 98 14:59:44
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5 ð3 98 15:ð4:36
Average Response Time (Seconds)
Itv
End
ð 1.ðð 2.ðð 3.ðð 4.ðð 5.ðð 6.ðð 7.ðð
----- +----+----+----+----+----+----+----+----+----+----+----+----+----+----+
\\\
15/ð5 RRRRRRRRRRRR
Response Components:
R = CPU + Disk + Wait-to-Ineligible
w = Exceptional Wait

Figure 7-35. Job Summary: Interactive Response Time by 5-Minute Intervals

Scatter Diagram of Interactive Transactions by 5-Minute

Intervals–Sample

Chapter 7. Performance Reports—Manager Feature 7-55

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Interactive Response Time by 5 Minute Intervals Page ðð88
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Average Response Time (Seconds)
| Itv
| End ð 1.ðð 2.ðð 3.ðð 4.ðð 5.ðð 6.ðð 7.ðð
| ----- +----+----+----+----+----+----+----+----+----+----+----+----+----+----+
| Response Components:
| R = CPU + Disk + Wait-to-Ineligible
| w = Exceptional Wait
| Job Summary Report 9/ð5/98 14:ð8:5ð
| Scatter Diagram of Interactive Transactions by 5 Minute Intervals Page ðð89
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Response Time vs Number of Transactions per Hour
| ------------------------------------------------
| OVFLW
| -----
| 7.ðð+
| -
| R -
| E -
| S 6.ðð+
| P -
| O -
| N -
| S 5.ðð+
| E -
| -
| T -
| I 4.ðð+
| M -
| E -
| -
| I 3.ðð+
| N -
| -
| S -
| E 2.ðð+
| C -
| O -
| N -
| D 1.ðð+
| S -
| -
| - ð
| ð.ðð+----+----+----+----+----+----+----+----+----+----+----+----+----+----+:O
| ð 1ðð 2ðð 3ðð 4ðð 5ðð 6ðð 7ðð:F
| Number of Transactions per Hour
| Legend: 1-9 Indicates the Number of Occurrences.
| \ Indicates more than 9 Occurrences.
| ð Identifies Average of All Occurrences.

Figure 7-36. Job Summary: Interactive Transactions by 5-Minute Intervals

Interactive Program Statistics–Sample

7-56 Performance Tools V4R2

 Job Summary Report 5/ð7/98 13:52:1ð
Interactive Program Statistics Page ðð19
Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 98 14:59:44
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5 ð3 98 15:ð4:36

Cum ----- Sync Disk I/O Rqs/Tns ----- Async Short Seize Cum
Number Program CPU CPU CPU DB DB NDB NDB DIO Rsp Wait Wait Pct Pct
Rank Tns Name /Tns Util Util Read Write Read Write Sum /Tns /Tns /Tns /Tns Tns Tns
---- ------ ---------- ------ ---- ---- ----- ----- ----- ----- ----- ----- ------ ------ ------ ---- ----
1 147 QUIINMGR .ð85 4.3 4.3 1 4 11 15 1ð .792 .ð31 65.3 65.3
2 32 QSPDSPF .ðð7 .1 4.3 1 1 1 .ð47 14.2 79.6
3 19 QPTPRCSS .ð23 .2 4.5 1 1 .ð51 8.4 88.ð
4 17 QUYLIST .ð63 .4 4.9 11 2 13 2 .411 7.6 95.6
5 3 QSUBLDS .1ð1 .1 5.ð 32 32 1.ð21 1.3 96.9
6 2 QUOCPP .ð34 5.ð 6 5 11 2 .433 .ð35 .9 97.8
7 2 QUIALIST .ð13 5.ð 1 1 .ð34 .9 98.7
8 1 \TRACEOFF\ 9.5ð8 3.3 8.2 27 2ð9 1852 257ð 4658 2118 157.268 .ð39 .4 99.1
9 1 QMHDSMSS .ð62 8.3 3 3 .135 .4 99.6
1ð 1 QUOCMD .ð44 8.3 1 1 .ð68 .4 1ðð.ð

Figure 7-37. Job Summary: Interactive Program Statistics

Summary of Seize/Lock Conflicts by Object–Sample

Job Summary Report 5/ð7/968 13:52:1ð
Summary of Seize/Lock Conflicts by Object Page ðð32
Member . . . : MON3D7CRT Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . :ð5 13 98 11:14:15
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5 13 98 12:14:ð1
-------- Interactive Waiters --------- ------- Non-Interactive Waiters -------
----- Locks ----- ---- Seizes ----- ----- Locks ----- ----- Seizes -----
Type Library File Member Number Avg Sec Number Avg Sec Number Avg Sec Number Avg Sec
------ ---------- ---------- ---------- ------ -------- ------ -------- ------ -------- ------ --------
DS CVTV3R2CAJ QAPMJOBS 3 .ð8ð
DS CVTV3R2CAJ QAPMLIOP 2 .ðð1
DS CVTV3R2CAJ QAPMPOOL 2 .1ð6
DS CVTV3R2CAJ QAPMRESP 2 .ð87
DS QUSRSYS QASNADSQ 1 .4ð6
DSI CVTV3R2CAJ QAPMCONF 2 .ðð6
DSI CVTV3R2CAJ QAPMLIOP 2 .ð13
DSI CVTV3R2CAJ QAPMPOOL 1 .ð15
FILE QSPL Qð4ð79Nðð3 14 .428
JOBQ QSYS QNMSVQ 3 .ð17 1 .ð62
JOBQ QSYS QSYSNOMAX 8 .ð2ð
LIB QRECOVERY 2 .ð92
LIB QSPL 8 .ð46
LIB QSVMSS 14 .ð38
LIB QUSRSYS 8 .197
LIB SOFIACN 1
MI Q QUSRSYS QS2RRAPPN 2 1.263
MSGQ QSYS QHST 7 .ð38 8 .343
OUTQ QUSRSYS QEZJOBLOG 6 .ð21
SMIDX QSVMSS QCQJMSMI 2
SPLCB QSPSCB 6 2.556
USRPRF MORIHE 4 .ð71
USRPRF QDBSHR 22 .ð39
USRPRF QSVCCS 21 .ð43
USRPRF QSYS 1 .ð38
1Eð1ð1 1 .ð29
---------------------------------------------- ------ -------- ------ -------- ------ -------- ------ --------
\ Total Conflicts and Avg Sec/Conflict 36 .847 191 .ð65
------ ------
\ Total Transactions With Conflicts
------ ------
\ Averages Per Conflict Transaction

Figure 7-38. Job Summary: Summary of Seize/Lock Conflicts by Object

Chapter 7. Performance Reports—Manager Feature 7-57

 Priority-Jobtype-Pool Statistics–Sample
| Job Summary Report 9/ð5/98 14:ð8:5ð
| Priority-Jobtype-Pool Statistics Page ðð9ð
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Job CPU --- Disk I/O Requests ---- Number
| Pty Type Pool Seconds Sync Async Tns
| --- ---- ---- ----------- ----------- ----------- ----------
| ðð B ð2 11.ð98 11ð4 358
| ðð L ð1 51.588 675 25569
| ðð L ð2 .156 14
| ðð M ð2 .ð22 26
| ðð S ð2 2.49ð 112
| ð9 S ð2
| 1ð B ð2
| 16 S ð2
| 2ð ð4 82.187 1321 27
| 2ð A ð2 .227 84 2ð
| 2ð B ð2 .658 11ð
| 2ð B ð4 .297 3
| 2ð BJ ð2
| 2ð I ð4 1.337 474 1 46
| 2ð L ð1
| 2ð S ð2
| 25 B ð2 .ð38 23
| 25 BJ ð2
| 3ð B ð2 .ð31 19
| 35 B ð2 .ð23 2ð
| 36 L ð1
| 4ð A ð2 .ðð6 7
| 4ð B ð2 .ð71 57
| 4ð X ð2
| 49 L ð1
| 5ð A ð2 .ðð7 6
| 5ð B ð2 .118 74
| 52 L ð1 .ð29 1ð 24
| 52 S ð2
| 6ð S ð2
| 68 L ð1 2.32ð 2
| 84 L ð1 .396 17ð
| 98 L ð1 .ðð9 1

Figure 7-39. Job Summary: Priority-Jobtype-Pool Statistics

Job Statistics–Sample
| Job Summary Report 9/ð5/98 14:ð8:5ð
| Job Statistics Page ðð91
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| JOBS WITH MOST TRANSACTIONS
| T P Cum Sync Async Number Cum
| Job User Name/ Job y t Nbr Rsp CPU CPU CPU DIO DIO Nbr Nbr Nbr Conflict Pct Pct
| Rank Name Thread Number Pl p y Tns /Tns /Tns Util Util /Tns /Tns W-I A-I Evt Lck Sze Tns Tns
| ---- ---------- ---------- ------ -- -- -- ----- ------ ------ ---- ---- ----- ----- --- --- --- --- --- ---- ----
| 1 QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð 43 .ð35 .ð18 .2 .2 93.5 93.5
| 2 QPADEVðð26 SOLBERG ð13841 ð4 I 2ð 3 4.918 .179 .2 .4 154 6.5 1ðð.ð
| 3
| 4
| 5
| 6
| 7
| 8
| 9
| 1ð
| JOBS WITH LARGEST AVERAGE RESPONSE TIME

Figure 7-40. Job Summary: Job Statistics

7-58 Performance Tools V4R2

 Interactive Program Statistics–Sample
Job Summary Report 5/ð7/98 13:52:1ð
Interactive Program Statistics Page ðð2ð
Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 98 14:59:44
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5 ð3 98 15:ð4:36
PROGRAMS WITH HIGHEST CPU/TNS
Cum ----- Sync Disk I/O Rqs/Tns ----- Async Short Seize Cum
Number Program CPU CPU CPU DB DB NDB NDB DIO Rsp Wait Wait Pct Pct
Rank Tns Name /Tns Util Util Read Write Read Write Sum /Tns /Tns /Tns /Tns Tns Tns
---- ------ ---------- ------ ---- ---- ----- ----- ----- ----- ----- ----- ------ ------ ------ ---- ----
1 1 \TRACEOFF\ 1 4658 2118 157.268 .ð39 .4 .4
2 1
3 1
4 1
5 1
6 1
7 1
8 1
9 1
1ð 1

Figure 7-41. Job Summary: Interactive Program Statistics

Individual Transaction Statistics–Sample

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Individual Transaction Statistics Page ð1ð1
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| TRANSACTIONS WITH LONGEST EXCESSIVE ACTIVITY LEVEL WAIT TIME
| Rank Value Time Program Job Name User Name Number Thread Pool Type Priority
| ---- ------------ ------------ ---------- ---------- ---------- ------ -------- ---- ---- --------
| 1 .ð18 12.48.15.8ð3 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 2 .ðð5 12.48.ð2.152 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 3
| 4
| 5
| 6
| 7
| 8
| 9
| 1ð
| TRANSACTIONS WITH LONGEST ACTIVE TIME
| Rank Value Time Program Job Name User Name Number Thread Pool Type Priority
| ---- ------------ ------------ ---------- ---------- ---------- ------ -------- ---- ---- --------
| 1 7.566 12.51.43.519 QUIINMGR QPADEVðð26 SOLBERG ð13841 ð4 I 2ð
| 2 6.536 12.51.28.115 QUIINMGR QPADEVðð26 SOLBERG ð13841 ð4 I 2ð
| 3 .642 12.51.31.29ð QUIINMGR QPADEVðð26 SOLBERG ð13841 ð4 I 2ð
| 4 .119 12.48.1ð.549 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 5 .ð84 12.47.49.115 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 6 .ð71 12.48.ð6.846 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 7 .ð48 12.47.59.476 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 8 .ð46 12.47.56.382 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 9 .ð46 12.47.44.114 QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð
| 1ð .ð38 12.48.16.17ð QUIINMGR QPADEVððð9 SUSTAITA ð13832 ð4 I 2ð

Figure 7-42. Job Summary: Individual Transaction Statistics

Longest Seize/Lock Conflicts–Sample

Chapter 7. Performance Reports—Manager Feature 7-59

| Job Summary Report 11/11/98 8:35:42
| Longest Seize/Lock Conflicts Page ðð31
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . :11/11/98 ð7:58:1ð
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :11/11/98 ð8:18:45
| Job User Name/ Job Holder- Job Name.. User Name. Number Pool Type Pty
| Rank Value Time Name Thread Number Pl Typ Pty S/L Object- Type.. Library... File...... Member.... RRN......
| ---- ------------ ------------ ---------- ---------- ------ -- --- --- --- ---------------------------------------------------------
| 1 2ð.679 ð8.ðð.43.582 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP CSTFIL ððððð2ððð
| 2 15.999 ð8.ðð.ð9.324 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP CSTFIL ððððð1ððð
| 3 14.183 ð8.ð1.16.8ð7 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP CSTFIL ððððð3ððð
| 4 .ð34 ð8.ðð.25.331 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP ITMFIL ððððð1ððð
| 5 .ð23 ð8.ð1.ð4.268 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP ITMFIL ððððð2ððð
| 6 .ð22 ð8.ð1.3ð.999 QPADEVðð17 ðððððððD ð23398 ð4 I ð1 L HOLDER- QPADEVðð16 COOK ð23399 ð4 I 2ð
| OBJECT- DS PFREXP ITMFIL ððððð3ððð

Figure 7-43. Job Summary: Longest Seize/Lock Conflicts

Longest Holders of Seize/Lock Conflicts–Sample

| Job Summary Report 11/11/98 8:35:42
| Longest Holders of Seize/Lock Conflicts Page ðð32
| STD I, B, then java/threads
| Member . . . : Q98315ð757 Model/Serial . : 51ð-2144/1ð-16CAD Main storage . . : 512.ð M Started . . . . :11/11/98 ð7:58:1ð
| Library . . : PFRRES42 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :11/11/98 ð8:18:45
| Job User Name/ Job ---------------------- Object -----------------------
| Rank Value Time Name Thread Number Pl Typ Pty S/L Type Library File Member RRN
| ---- -------- ------------ ---------- ---------- ------ -- --- --- --- ------ ---------- ---------- ---------- ---------
| 1 2ð.679 ð8.ðð.43.581 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP CSTFIL ððððð2ððð
| 2 15.999 ð8.ðð.ð9.324 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP CSTFIL ððððð1ððð
| 3 14.183 ð8.ð1.16.8ð8 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP CSTFIL ððððð3ððð
| 4 .ð34 ð8.ðð.25.332 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP ITMFIL ððððð1ððð
| 5 .ð23 ð8.ð1.ð4.269 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP ITMFIL ððððð2ððð
| 6 .ð22 ð8.ð1.3ð.999 QPADEVðð16 ðððððððD ð23399 ð4 I 2ð L DS PFREXP ITMFIL ððððð3ððð

Figure 7-44. Job Summary: Longest Holders of Seize/Lock Conflicts

Batch Job Analysis–Sample

| Job Summary Report 9/ð5/98 14:ð8:5ð
| Batch Job Analysis Page ð1ð6
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| T P Sync Async --- Synchronous --- Excp
| Job User Name/ Job y t Elapsed CPU CPU Disk Disk BCPU --DIO/Sec-- Wait
| Name Thread Number Pl p y Start Stop Seconds Seconds Util I/O I/O /DIO Elp Act Ded Sec
| ---------- ---------- ------ -- -- -- -------- -------- ---------- ---------- ----- -------- -------- -------- --- --- --- ---------
| QPASVRP QSYS ð13724 ð2 B 5ð 12.47.42 12.53.25 342.791 .ðð8 9 .ððð9 29 2.56
| QECS QSVSM ð13727 ð2 B 5ð 12.47.42 12.53.25 342.773 .ð58 21 .ðð28 1 28 319.27
| QNSCRMON QSVSM ð1373ð ð2 B 5ð 12.47.42 12.53.25 342.773 .ðð7 8 .ððð9 29 2.54
| QSYSSCD QPGMR ð13731 ð2 B 1ð 12.47.42 12.53.25 342.754 342.75
| QZDAINIT QUSER ð13733 ð2 BJ 2ð 12.47.42 12.53.25 342.742 342.74
| QPASVRS QSYS ð13732 ð2 B 3ð 12.47.42 12.53.25 342.721 .ð12 5 .ðð24 28 .55
| QPWFSERVSO QUSER ð13735 ð2 BJ 2ð 12.47.42 12.53.25 342.7ðð 342.7ð
| QPWFSERVS2 QUSER ð13736 ð2 BJ 2ð 12.47.42 12.53.25 342.685 342.68
| QPWFSERV QUSER ð13739 ð2 BJ 2ð 12.47.42 12.53.25 342.653 342.65
| QPWFSERV QUSER ð1374ð ð2 BJ 2ð 12.47.42 12.53.25 342.635 342.63
| QPWFSERV QUSER ð13742 ð2 BJ 2ð 12.47.42 12.53.25 342.618 342.61
| QZLSFILE QUSER ð13743 ð2 BJ 2ð 12.47.42 12.53.25 342.599 342.59
| QPASVRS QSYS ð13737 ð2 B 3ð 12.47.42 12.53.25 342.585 .ðð6 5 .ðð12 29 .68
| QNMAPINGD QUSER ð13746 ð2 BJ 25 12.47.42 12.53.25 342.561 342.56
| QNMAREXECD QUSER ð13748 ð2 BJ 25 12.47.42 12.53.25 342.522 342.52
| QPASVRS QSYS ð13747 ð2 B 3ð 12.47.42 12.53.25 342.5ð2 .ðð9 5 .ðð18 29 .43
| .
| .
| .

Figure 7-45. Job Summary: Batch Job Analysis

7-60 Performance Tools V4R2

| Concurrent Batch Job Statistics–Sample
| Job Summary Report 9/ð5/98 14:ð8:5ð
| Concurrent Batch Job Statistics Page ð1ð8
| Transaction Report Summary
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Job Number Elapsed CPU Excp Sync Async
| Set Pty Jobs Seconds Seconds Wait Disk I/O Disk I/O
| ------ --- ------ -------- -------- -------- -------- --------
| 1 ðð 1 341.352 11.ð98 291.49ð 11ð4 358
| 2 1ð 1 342.754 342.75ð
| 3 2ð 1 .467 .ð11 .44ð
| 4 2ð 1 .555 .ð1ð .54ð
| 5 2ð 1 .657 .ð11 .6ðð
| 6 2ð 1 .787 .ð6ð .6ðð 1
| 7 2ð 1 .877 .ð11 .86ð
| 8 2ð 1 .971 .ð12 .95ð
| 9 2ð 1 1.ð79 .ð15 1.ð4ð 1
| 1ð 2ð 1 1.167 .ð12 1.13ð 1
| 11 2ð 1 1.25ð .ð1ð 1.23ð
| 12 2ð 1 1.341 .ð1ð 1.32ð
| 13 2ð 1 1.433 .ð11 1.41ð
| 14 2ð 1 1.514 .ð12 1.49ð
| 15 2ð 1 1.6ð4 .ð13 1.58ð
| 16 2ð 1 1.7ð8 .ð1ð 1.69ð
| .
| .
| .

Figure 7-46. Job Summary: Concurrent Batch Job Statistics

Report Selection Criteria-Sample

Job Summary Report ð5/ð7/95 13:4ð:32
Report Selection Criteria Page ðð14
Sample Transaction Report

Member . . . : CAJð5ð3 Model/Serial . : 51ð-2144/1ð-ð8BCD Main storage . . : 384.ð M Started . . . . : ð5 ð3 96 14:59:44

Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release : 3/ 7.ð Stopped . . . . : ð5 ð3 96 15:ð4:36

Select Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - Jobname1
Jobname2
Jobnum

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Select parameters were chosen.

Omit Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Omit parameters were chosen.

Options Selected - SS Include Special Summary Reports

Figure 7-47. Job Summary Report: Report Selection Criteria

Chapter 7. Performance Reports—Manager Feature 7-61

 Transaction Report Option–Sample
| Note: This Transaction Report ran a collection with thread activity. The report
| header shows the thread identifier because the job is a secondary thread.
| Transaction Report 9/11/986:35:48
| Page ððð1
| Member . . . :
Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . :
THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Job name . . :
QTMSNMPRCV User name . . . : QTCP Job number . . . : ð13771 TDE/Pl/Pty/Prg . : ð185/ð2/5ð/YES
| Thread . . . :
ðððððððDF
| E T CPU ---- Physical I/O Counts ----- \\\\\ Transaction Response Time (Sec/Tns) \\\\\\ -BMPL-
| x y Sec ----- Synchronous ------ Async \\\\\\ - Activity Level Time - Inel Long C I Seize
| c Program p Per DB DB NDB NDB Disk \\\\ Short Seize Time Wait u n Hold Key/
| Time p Name e Tns Read Wrt Read Wrt Sum I/O \\ Active Wait Cft A-I/W-I Lck/Oth r l Time Think
| -------- - ---------- - ------- ---- ---- ---- ---- ---- ----- -------- ------- ------- ------- ------- ------- -- -- ----- ------
| 12.47.44 Y QSOSRV1 .ð11 5 5 ð .1ð9 .ð98 11 2.3
| -----------------------------------------------
| J O B S U M M A R Y D A T A (T O T A L S)
| -----------------------------------------------
| Average .ð11 ð ð 7 ð 7 ð .1ð9 .ð98 .ððð .ððð .ððð .ððð .ð 2.3
| Count 1 1 1
| Minimum .ð11 5 .1ð9 .ð98 2.3
| Maximum .ð11 5 .1ð9 .ð98 2.3
| Total/Job .ð11 5 ð 342.ð16 Elapsed .ð Percent CPU Utilization

Figure 7-48. Transaction Report

Transition Report Option–Sample

| Note: This Transition Report ran a collection with thread activity. The report
| header shows the thread identifier because the job is a secondary thread.
| Transition Report 9/11/986:34:3ð
| Page ððð1
| Member . . . :
Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . :ð5/22/98 12:47:42
| Library . . :
THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . :ð5/22/98 12:53:25
| Job name . . :
QTMSNMPRCV User name . . . : QTCP Job number . . . : ð13771 TDE/Pl/Pty/Prg . : ð185/ð2/5ð/YES
| Thread . . . :
ððððððð5
| Job type . .B : Elapsed Time -- Seconds Sync/Async Phy I/O -MPL-
| ----------------------- ------------------------ C I Last 4 Programs in Invocation Stack
| State Wait Long Active Inel CPU DB DB NDB NDB u n -------------------------------------------
| Time W A I Code Wait /Rsp\ Wait Sec Read Wrt Read Wrt Tot r l Last Second Third Fourth
| ------------ ----- ---- ------- ------- ------- -------- ---- ---- ---- ---- ---- -- -- ---------- ---------- ---------- ----------
| 12.47.42.587 \TRACE ON
| 12.47.44.9ð4 ->A 2.317 .ð11 .ð11 5 11
| PAG= 4 PWrt= ð
| 12.47.45.ðð1 W<- .ð98 11 QSOKERN3 QSOKERN3 QSOSRV1 QTOSRCVR
| ---------- QSOSRV1 .1ð9\ .ð11 ð ð 5 ð 5\ PAG= 4 PWrt= ð
| 12.53.24.6ð3 /OFF .ðð4 2 PAG= 1 PWrt= ð
| 12.53.24.6ð3 \TRACE OFF
| -----------------------------------------------
| J O B S U M M A R Y D A T A (T O T A L S)
| -----------------------------------------------
| CPU ---- Physical I/O Counts ----- \\\\\ Transaction Response Time (Sec/Tns) \\\\\\ -BMPL-
| Sec ----- Synchronous ------ Async \\\\\\ - Activity Level Time - Inel Long C I Seize
| Per DB DB NDB NDB Disk \\\\ Short Seize Time Wait u n Hold Key/
| Tns Read Wrt Read Wrt Sum I/O \\ Active Wait Cft A-I/W-I Lck/Oth r l Time Think
| ------- ---- ---- ---- ---- ---- ----- -------- ------- ------- ------- ------- ------- -- -- ----- ------
| Average .ð11 ð ð 7 ð 7 ð .1ð9 .ð98 .ððð .ððð .ððð .ððð .ð 2.3
| Count 1 1 1
| Minimum .ð11 5 .1ð9 .ð98 2.3
| Maximum .ð11 5 .1ð9 .ð98 2.3
| Total/Job .ð11 5 ð 342.ð16 Elapsed .ð Percent CPU Utilization

Figure 7-49. Transition Report

7-62 Performance Tools V4R2

Lock Report

Printing the Lock Report

Use the PRTLCKRPT command. The PRTLCKRPT command uses trace output
from the STRPFRMON command, so the STRPFRMON command must be run
first.

Collecting Trace Data

To create the trace data:
1. Start the performance monitor (use the STRPFRMON command) and specify
tracing (TRACE(*ALL)).
2. When the monitor job completes, print the report. The value for the MBR
parameter on the print command should be the same as the value for the LIB
parameter on the print command.
Note: You can end the performance monitor before the specified interval has
ended by using the End Performance Monitor (ENDPFRMON)
command.

When you use the PRTLCKRPT command, the following file is used as input:
File Description
QAPMDMPT Database file that is created by the STRPFRMON command and
updated by the PRTTNSRPT command.

See “QTRDMPT File” on page 7-96 for a description of the database file.

Following are the output files from the PRTLCKRPT command:

File Description
QPPTLCK Printer file
QAPTLCKD Database file

See “QAPTLCKD File” on page 7-101 for the database file description.
Note: In the following description, the term lock means lock or seize unless other-
wise noted.

The PRTLCKRPT command produces several report formats. An optional detail list
of the resource management trace records from QAPMDMPT prints first. This list
may be sorted by the times that a lock occurred, the name of the job requesting the
lock, the name of the job holding the lock, or the name of the locked object. The list
may print four times (once for each of these sequences).

Consider the following points when you use the PRTLCKRPT command:
Ÿ The PRTTNSRPT output may show a high incidence of wait-to-ineligible state
transitions in the transaction summary output. If this situation occurs, it could
mean that many jobs are waiting for internal system object locks and holding
an activity level while waiting. The PRTLCKRPT report may identify these locks.
Ÿ The Detailed Lock Conflicts Report (shown in Figure 7-51 on page 7-66)
shows each object lock conflict that meets the specified selection values. Do

Chapter 7. Performance Reports—Manager Feature 7-63

 not assume that each conflict shown for an object lock is associated with a
separate request for the object from the program that originally requested it.

When multiple requests (from multiple jobs) cause contention for an object, the
requests are processed in the order received, by job priority. When conflicts
occur, multiple lock requests are made by internal programs in behalf of the
program that originally made the request, until the lock is granted. These
internal requests appear on the summary, resulting in more conflicts than actu-
ally occurred from the originating program’s viewpoint.

PRTLCKRPT processing does not analyze the internal lock conflicts and relate
them to the original request.

What Is the Lock Report?

The Lock Report provides information about lock and seize conflicts during system
operation. With this information you can determine if jobs are being delayed during
processing because of unsatisfied lock requests or internal machine seizes. These
conditions are also called waits. If they are occurring, you can determine which
objects the jobs are waiting for and the length of the wait.

Next, these summaries print detail listings summarized by:

Ÿ Requesting job
Ÿ Holding job
Ÿ Object name

Figure 7-51 on page 7-66 shows a sample of the detail listing, sorted by time of
day (in this case). The report options were selected to include only locks lasting at
least two seconds that occurred between 13:33:00 and 13:34:00 (as noted in the
footer printed at the bottom of the detail page).

Figure 7-52 on page 7-66 shows a sample of the Requesting Job Summary
section of the same report. The other summary sections have a similar format.

See “Sample Lock Reports” on page 7-65.

See “Performance Report Columns” on page 7-102 for definitions of specific

columns in the reports.

Analyzing Seize/Lock Conflicts

Seizes/locks are system-locking functions that ensure integrity during certain oper-
ations. For example, the system uses a seize during logical file maintenance when
the underlying physical files are changed.

Conflicts occur when one job has an object lock or seize and another job requests
control of the same object. A common example of a lock conflict is when a job
reads a record for update and a second job requests a lock for the same record.

If the Print Transaction Report (PRTTNSRPT) job summary output shows a high
number for either the number of lock or the number of seize conflicts, look at the
Transaction Detail Report and Transition Detail Report to further analyze the situ-

7-64 Performance Tools V4R2

 ation. You can also use the PRTLCKRPT command to print the Seize/Lock Conflict
Report to see what conflicts occurred.

If the PRTTNSRPT command output shows several lock waits, or system

throughput is low and the processing unit time and disk use is also low, these con-
ditions could be caused by lock-wait conflicts occurring in jobs due to contention for
files, records, or other objects. Analyze the resource management trace data using
the PRTLCKRPT command to determine a cause.

You can normally expect to see some conflicts occur for a short period of time on
some objects. If you see several lock conflicts occur for nondatabase objects, it
may be a normal situation (such as writers and jobs contending for output queues).
However, if the locks last a long time (more than 5 to 10 seconds), and they cause
objectionable delays to end users, this situation could indicate that you need to
make some changes to the operational environment.

If the report shows several database record locks that last for more than 5 to 10
seconds, a program may have read a record for update and continued processing
without releasing (writing) the record. This situation is normal in many applications.
However, in a heavily loaded system, the job that holds the record lock may reach
the end of its time slice while it holds the lock. When this condition occurs, it delays
other jobs that need the record.

If the report shows several seizes that last for a period of time (over 1 second), this
condition can indicate object contention problems. To ensure the accuracy of the
object, the system does not allow access to the object until all the necessary
changes are made.

| Thread Data
| As you will see in the Lock Report examples, if the data collection contains thread
| activity, and if the job is a secondary thread, the detail in the report shows the job
| name/thread identifier/job number value. If there is no thread activity, the detail
| shows the job name/user name/number value. Figure 7-50 shows a comparison
| between a job that is a secondary thread and a job that is not a secondary thread.
| 9/24/98 7:4ð:ð8 Seize/Lock Wait Statistics by Time of Day Page 1
| TOD of Length Object Record
| Wait of Wait L Requestor's Job Name Holder's Job Name Type Object Name Number
| -------- ------- - ---------------------------- ---------------------------- ------ -------------------------------- ----------
| 13.ð1.28 179 MNTASK QPADEVððð9 SUSTAITA ð13917 LIB QUSRSYS
| 13.ð4.4ð 2ð TPCRTMAX ðððððð57 ð13922 TPCRTMAX SUSTAITA ð13923 LIB QUSRSYS
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 ð ms minimum wait
| 9/24/98 7:4ð:ð8 Seize/Lock Wait Statistics Summary Page 2
| Locks Seizes
| Requestor's Job Name Count Avg Length Count Avg Length
| ------------------------------------- ------------- ------------- ------------- -------------
| MNTASK 2 1ð4
| TPCRTMAX SUSTAITA ð13922 ððððððB8 1 193
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 ð ms minimum wait

| Figure 7-50. Example of a Detail Listing with Thread Data

Sample Lock Reports

See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Chapter 7. Performance Reports—Manager Feature 7-65

 Lock Report–Detail
| 9/24/98 7:4ð:ð8 Seize/Lock Wait Statistics by Time of Day Page 1
| TOD of Length Object Record
| Wait of Wait L Requestor's Job Name Holder's Job Name Type Object Name Number
| -------- ------- - ---------------------------- ---------------------------- ------ -------------------------------- ----------
| 13.ðð.43 29 MNTASK TPCRTMAX SUSTAITA ð13921 LIB QUSRSYS
| 13.ð1.28 179 MNTASK QPADEVððð9 SUSTAITA ð13917 LIB QUSRSYS
| 13.ð4.4ð 2ð TPCRTMAX ðððððð57 ð13922 TPCRTMAX SUSTAITA ð13923 LIB QUSRSYS
| 13.ð5.ð8 1ð TPCRTMAX ðððððð8C ð13922 TPCRTMAX SUSTAITA ð13924 LIB QUSRSYS
| 13.ð5.38 4 TPCRTMAX ðððððð5C ð13923 TPCRTMAX SUSTAITA ð13925 LIB QUSRSYS
| 13.ð5.57 2ðð TPCRTMAX ðððððð26 ð13924 TPCRTMAX SUSTAITA ð13925 LIB QUSRSYS
| 13.ð5.57 193 TPCRTMAX ððððððB8 ð13922 TPCRTMAX SUSTAITA ð13925 LIB QUSRSYS
| 13.ð7.58 1ðð TPCRTMAX ðððððð5B ð13923 TPCRTMAX SUSTAITA ð13925 LIB QUSRSYS
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 ð ms minimum wait
| 9/24/98 7:4ð:ð8 Seize/Lock Wait Statistics Summary Page 2
| Locks Seizes
| Requestor's Job Name Count Avg Length Count Avg Length
| ------------------------------------- ------------- ------------- ------------- -------------
| MNTASK 2 1ð4
| TPCRTMAX SUSTAITA ð13922 ððððððB8 1 193
| TPCRTMAX SUSTAITA ð13922 ðððððð57 1 2ð
| TPCRTMAX SUSTAITA ð13922 ðððððð8C 1 1ð
| TPCRTMAX SUSTAITA ð13923 ðððððð5B 1 1ðð
| TPCRTMAX SUSTAITA ð13923 ðððððð5C 1 4
| TPCRTMAX SUSTAITA ð13924 ðððððð26 1 2ðð
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 ð ms minimum wait
| 9/24/98 7:4ð:ð8 Seize/Lock Wait Statistics Summary Page 3
| Locks Seizes
| Holder's Job Name Count Avg Length Count Avg Length
| ------------------------------------- ------------- ------------- ------------- -------------
| QPADEVðððB SUSTAITA ð79289 6 77,39ð
| QPADEVðððC SUSTAITA ð7929ð 3 129,ð22
| QPADEVðð15 SUSTAITA ð79291 3 114,954
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 5ðð ms minimum wait

Figure 7-51. Example of a Detail Listing

Lock Report–Summary
| 9/24/98 7:39:48 Seize/Lock Wait Statistics Summary Page 1
| Locks Seizes
| Requestor's Job Name Count Avg Length Count Avg Length
| ------------------------------------- ------------- ------------- ------------- -------------
| QPADEVðððC SUSTAITA ð7929ð 4 9ð,394
| QPADEVðð15 SUSTAITA ð79291 4 117,47ð
| QPADEVðð25 SUSTAITA ð79139 4 91,2ð3
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 5ðð ms minimum wait
| 9/24/98 7:39:48 Seize/Lock Wait Statistics Summary Page 2
| Locks Seizes
| Holder's Job Name Count Avg Length Count Avg Length
| ------------------------------------- ------------- ------------- ------------- -------------
| QPADEVðððB SUSTAITA ð79289 6 77,39ð
| QPADEVðððC SUSTAITA ð7929ð 3 129,ð22
| QPADEVðð15 SUSTAITA ð79291 3 114,954
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 5ðð ms minimum wait
| 9/24/98 7:39:48 Seize/Lock Wait Statistics Summary Page 3
| Object Locks Seizes
| Type Object Name Count Avg Length Count Avg Length
| ------ -------------------------------- ------------- ------------- ------------- -------------
| DS QREXSRC SUSTAITA CRTETH 5 69,ð95
| MQLOCK QMH MESSAG E QUEUE LO 3 123,276
| PGM DMPSPC SUSTAITA 4 12ð,24ð
| Member LOCKDATA Library RWSDATA Period from ðð.ðð.ðð through 23.59.59 5ðð ms minimum wait

Figure 7-52. Example of Summary by Requesting Job

Job Interval Report

Printing the Job Interval Report

Use the Print Job Report (PRTJOBRPT) command, or select option 5 (Job report)
on the Print Performance Reports display.

7-66 Performance Tools V4R2

What Is the Job Interval Report?
This report, like other similar reports, is produced from the sample data collected
with the STRPFRMON command. The four major sections of this report show infor-
mation on all or selected intervals and jobs, including detail and summary informa-
tion for interactive jobs and for noninteractive jobs. Because the report can be long,
you may want to limit the output by selecting the intervals and jobs you want to
include. For example, you might want to specify OMTSYSTSK(*YES) on the
PRTJOBPRT command to print only the user jobs and omit the system tasks. Or,
you can specify OMTSYSTSK(*NO) and include the system tasks.

If a value is too large to fit in the allotted space, a 9 is printed in each numeric field
in the report.

The following are sections of the Job Interval Report:

Ÿ Interactive Job Summary
Ÿ Noninteractive Job Summary
Ÿ Interactive Job Detail
Ÿ Noninteractive Job Detail
Ÿ Report Selection Criteria

Interactive Job Summary

The Interactive Job Summary section of the Job Interval Report lists one line for all
selected interactive jobs that existed during each selected interval (a total of one
line per interval).

The information included in this section includes only valid interactive jobs with
CPU activity different than zero, or with any I/O activity.

See the sample report shown in Figure 7-53 on page 7-68.

Noninteractive Job Summary

The Noninteractive Job Summary section of the Job Interval Report lists one line
for all selected noninteractive jobs that existed during each selected interval (a total
of one line per interval).

The information included in this section includes only valid non-interactive jobs with
CPU activity different than zero, or with any I/O activity.

See the sample report shown in Figure 7-54 on page 7-69.

Interactive Job Detail

The Interactive Job Detail section of the Job Interval Report gives detailed informa-
tion by interval and job. One line is printed for each selected interactive job that
existed during each selected interval (generally more than one line per interval).

See the sample report shown in Figure 7-55 on page 7-70.

Chapter 7. Performance Reports—Manager Feature 7-67

 Noninteractive Job Detail
The Noninteractive Job Detail section of the Job Interval Report gives detailed infor-
mation by interval and job. One line is printed for each selected noninteractive job
that existed during each selected interval (generally more than one line per
interval).

See the sample report shown in Figure 7-56 on page 7-71.

Report Selection Criteria

The Report Selection Criteria section of the Job Interval Report gives the selection
values you chose to produce the report.

See the sample report shown in Figure 7-57 on page 7-72.

Sample Job Interval Reports

See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Interactive Job Summary–Sample

| Job Interval Report ð9/15/98 1ð:51:36
| Interactive Job Summary Page 1
| Sample Job Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| Itv Act Tns Rsp/ ------------------ Number of I/O -------------------- Tns/ CPU PAG Perm Arith
| End Jobs Count Tns DDM Sync Async Logical Cmn Hour Util Fault Write Ovrfl
| ----- ------ ---------- ------ --------- --------- --------- --------- --------- ---------- ---- ------ ------- ------
| ð9:46 2 34 2.41 ð 4ð73 399 188 ð 349 1.7 1ð 57ð ð
| 1ð:18 1 17 .41 ð 4ð1 62 63 ð 175 .3 1 1ð3 ð
| 1ð:33 2 24 1.ð8 ð 1247 151 159 ð 95 .3 3 85 ð
| 1ð:48 1 2 .ðð ð 43 4 ð ð 17 .ð ð 6 ð
| 11:ðð 1 3 .ðð ð 56 5 ð ð 981 1.8 1 1ð ð
| 11:46 1 7 .14 ð 23 3 ð ð 354 .3 1 11 ð
| 12:47 2 29 1.31 ð 2272 14 ð ð 1,11ð 5.9 11ð 25 ð
| 13:46 1 7 .14 ð 24 3 ð ð 354 .3 2 1ð ð
| 14:47 1 7 .14 ð 24 3 ð ð 35ð .3 ð 1ð ð
| 15:ð3 1 11 .9ð ð 513 22 4 ð 16ð .3 1 48 ð
| 15:18 2 9 2.11 ð 1425 89 47 ð 35 .1 12 1ð5 ð
| 15:33 2 6 2.16 ð 66ð 15ð 2ð8 ð 23 .1 11 183 ð
| 15:48 2 1ð .6ð ð 548 26 ð ð 39 .ð 16 35 ð
| 15:49 2 7 .28 ð 1ð6 6 ð ð 84ð .9 8 11 ð
| 16:24 1 ð .ðð ð 195 9 ð ð ð .ð 19 2ð ð
| 16:46 1 7 .14 ð 24 3 ð ð 35ð .3 1 11 ð
| 17:46 1 7 .14 ð 23 3 ð ð 354 .3 1 1ð ð
| 18:46 1 7 .14 ð 26 3 ð ð 354 .3 1 1ð ð
| 19:46 1 7 .14 ð 26 3 ð ð 35ð .3 2 1ð ð
| 2ð:46 1 7 .14 ð 23 3 ð ð 35ð .3 1 1ð ð
| 21:46 1 7 .14 ð 24 3 ð ð 354 .3 1 1ð ð
| 22:46 1 7 .ðð ð 25 3 ð ð 35ð .3 1 11 ð
| 23:46 1 7 .14 ð 48 9 ð ð 35ð .4 8 9 ð
| Itv End -- Interval end time (hour and minute)
| Act Jobs -- Number of active jobs in the interval
| Tns Count -- Number of transactions
| Rsp/Tns -- Average response time (seconds)
| DDM -- Number of logical DB I/O operations for DDM server jobs
| Sync -- Number of synchronous disk I/O operations
| Async -- Number of asynchronous disk I/O operations
| Logical -- Number of logical disk I/O operations
| Cmn -- Number of communications I/O operations
| Tns/Hour -- Average number of transactions per hour
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| PAG Fault -- Number of faults involving the Process Access Group
| Perm Write -- Number of permanent writes
| Arith Ovrfl -- Number of arithmetic overflow exceptions

Figure 7-53. Interactive Job Summary Section

7-68 Performance Tools V4R2

 Noninteractive Job Summary–Sample
| Job Interval Report ð9/15/98 1ð:51:36
| Non-Interactive Job Summary Page 2
| Sample Job Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| Itv Act CPU ---- Number of I/O Per Second ---- -- CPU/ I/O --- Line Page PAG Perm Arith
| End Jobs Util Sync Async Logical Cmn Sync Async Count Count Fault Write Ovrflw
| ----- ------- ---- ------- ------- ------- ------- ------- ------- --------- ------- ------- --------- -------
| ðð:31 34 31.8 3.8 .7 .1 .ð 82 428 2,611 46 1 1,1ð2 ð
| ðð:46 38 31.9 5.8 1.2 .2 .ð 55 258 2,619 5ð 4 1,954 ð
| ð1:ð1 41 31.7 5.6 1.ð .6 .ð 56 314 2,734 51 22 1,714 ð
| ð1:16 38 31.2 5.3 .8 .2 .ð 58 369 2,553 46 4 1,685 ð
| ð1:31 4ð 31.5 4.4 .8 .2 .ð 71 39ð 3,ð76 52 5 1,385 ð
| ð1:46 48 31.5 2.8 .7 .2 .ð 11ð 4ð7 3,728 68 63 1,113 ð
| ð2:ð1 47 31.6 1.7 .4 .2 .ð 184 666 3,924 75 4 664 ð
| ð2:16 42 32.ð 2.4 .5 .2 .ð 128 631 3,573 66 1ð 788 ð
| ð2:31 42 31.9 3.3 .7 .2 .ð 94 4ð2 3,228 61 1 1,14ð ð
| ð2:46 41 32.ð 2.ð .5 .2 .ð 156 563 2,916 57 4 777 ð
| ð3:ð1 37 31.7 5.7 1.ð .2 .ð 55 3ð9 2,776 52 2 1,741 ð
| ð3:16 37 32.ð 2.2 .4 .2 .ð 143 7ð3 2,771 51 ð 674 ð
| ð3:31 37 31.8 3.9 .8 .2 .ð 81 383 2,763 52 2 1,332 ð
| ð3:46 41 31.7 7.4 1.2 .2 .ð 42 257 2,8ð1 5ð 9 2,5ðð ð
| ð4:ð2 45 32.1 3.ð .5 .2 .ð 1ð6 568 3,47ð 61 9 1,ð23 ð
| ð4:17 46 32.ð 3.ð .6 .2 .ð 1ð6 483 4,ð15 72 18 1,ðð5 ð
| ð4:32 46 32.1 3.7 .9 .2 .ð 84 35ð 4,ð62 74 1ð 1,316 ð
| ð4:47 46 31.9 3.8 .8 .2 .ð 83 366 3,7ð5 72 6 1,349 ð
| ð5:ð2 4ð 32.1 .7 .1 .2 .ð 433 2,322 3,31ð 63 ð 279 ð
| ð5:17 41 32.ð 2.4 .5 .2 .ð 131 636 2,954 55 8 771 ð
| ð5:32 38 32.ð 3.9 1.ð .2 .ð 81 293 2,949 55 5 1,268 ð
| ð5:47 38 32.ð 1.8 .4 .2 .ð 172 771 2,895 53 5 688 ð
| ð6:ð2 41 31.8 3.2 .6 .2 .ð 98 458 2,9ðð 54 1 996 ð
| ð6:17 4ð 31.8 7.2 1.1 .2 .ð 44 284 2,956 54 2 2,382 ð
| ð6:32 48 32.1 3.3 .8 .2 .ð 95 375 3,732 67 12 1,13ð ð
| Itv End -- Interval end time (hour and minute)
| Act Jobs -- Number of jobs that were active during the interval
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| Sync I/O Per Second -- Average number of synchronous disk I/O operations per second
| Async I/O Per Second -- Average number of asynchronous disk I/O operations per second
| Logical I/O Per Second -- Average number of logical disk I/O operations per second
| Cmn I/O Per Second -- Average number of communications I/O operations per second
| CPU/ Sync I/O -- Avg number of CPU milliseconds per synchronous disk I/O operation
| CPU/ Async I/O -- Avg number of CPU milliseconds per asynchronous disk I/O operation
| Line Count -- Number of lines printed
| Page Count -- Number of pages printed
| PAG Fault -- Number of faults involving the Process Access Group
| Perm Write -- Number of permanent writes
| Arith Ovrflw -- Number of arithmetic overflow exceptions

Figure 7-54. Noninteractive Job Summary Section

Interactive Job Detail–Sample

Chapter 7. Performance Reports—Manager Feature 7-69

| Job Interval Report ð9/15/98 1ð:51:36
| Non-Interactive Job Summary Page 2
| Sample Job Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| Itv Act CPU ---- Number of I/O Per Second ---- -- CPU/ I/O --- Line Page PAG Perm Arith
| End Jobs Util Sync Async Logical Cmn Sync Async Count Count Fault Write Ovrflw
| ----- ------- ---- ------- ------- ------- ------- ------- ------- --------- ------- ------- --------- -------
| ðð:31 34 31.8 3.8 .7 .1 .ð 82 428 2,611 46 1 1,1ð2 ð
| ðð:46 38 31.9 5.8 1.2 .2 .ð 55 258 2,619 5ð 4 1,954 ð
| ð1:ð1 41 31.7 5.6 1.ð .6 .ð 56 314 2,734 51 22 1,714 ð
| ð1:16 38 31.2 5.3 .8 .2 .ð 58 369 2,553 46 4 1,685 ð
| ð1:31 4ð 31.5 4.4 .8 .2 .ð 71 39ð 3,ð76 52 5 1,385 ð
| ð1:46 48 31.5 2.8 .7 .2 .ð 11ð 4ð7 3,728 68 63 1,113 ð
| ð2:ð1 47 31.6 1.7 .4 .2 .ð 184 666 3,924 75 4 664 ð
| ð2:16 42 32.ð 2.4 .5 .2 .ð 128 631 3,573 66 1ð 788 ð
| ð2:31 42 31.9 3.3 .7 .2 .ð 94 4ð2 3,228 61 1 1,14ð ð
| ð2:46 41 32.ð 2.ð .5 .2 .ð 156 563 2,916 57 4 777 ð
| ð3:ð1 37 31.7 5.7 1.ð .2 .ð 55 3ð9 2,776 52 2 1,741 ð
| ð3:16 37 32.ð 2.2 .4 .2 .ð 143 7ð3 2,771 51 ð 674 ð
| ð3:31 37 31.8 3.9 .8 .2 .ð 81 383 2,763 52 2 1,332 ð
| ð3:46 41 31.7 7.4 1.2 .2 .ð 42 257 2,8ð1 5ð 9 2,5ðð ð
| ð4:ð2 45 32.1 3.ð .5 .2 .ð 1ð6 568 3,47ð 61 9 1,ð23 ð
| ð4:17 46 32.ð 3.ð .6 .2 .ð 1ð6 483 4,ð15 72 18 1,ðð5 ð
| ð4:32 46 32.1 3.7 .9 .2 .ð 84 35ð 4,ð62 74 1ð 1,316 ð
| ð4:47 46 31.9 3.8 .8 .2 .ð 83 366 3,7ð5 72 6 1,349 ð
| ð5:ð2 4ð 32.1 .7 .1 .2 .ð 433 2,322 3,31ð 63 ð 279 ð
| ð5:17 41 32.ð 2.4 .5 .2 .ð 131 636 2,954 55 8 771 ð
| ð5:32 38 32.ð 3.9 1.ð .2 .ð 81 293 2,949 55 5 1,268 ð
| ð5:47 38 32.ð 1.8 .4 .2 .ð 172 771 2,895 53 5 688 ð
| ð6:ð2 41 31.8 3.2 .6 .2 .ð 98 458 2,9ðð 54 1 996 ð
| ð6:17 4ð 31.8 7.2 1.1 .2 .ð 44 284 2,956 54 2 2,382 ð
| ð6:32 48 32.1 3.3 .8 .2 .ð 95 375 3,732 67 12 1,13ð ð
| Itv End -- Interval end time (hour and minute)
| Act Jobs -- Number of jobs that were active during the interval
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| Sync I/O Per Second -- Average number of synchronous disk I/O operations per second
| Async I/O Per Second -- Average number of asynchronous disk I/O operations per second
| Logical I/O Per Second -- Average number of logical disk I/O operations per second
| Cmn I/O Per Second -- Average number of communications I/O operations per second
| CPU/ Sync I/O -- Avg number of CPU milliseconds per synchronous disk I/O operation
| CPU/ Async I/O -- Avg number of CPU milliseconds per asynchronous disk I/O operation
| Line Count -- Number of lines printed
| Page Count -- Number of pages printed
| PAG Fault -- Number of faults involving the Process Access Group
| Perm Write -- Number of permanent writes
| Arith Ovrflw -- Number of arithmetic overflow exceptions

Figure 7-55. Job Interval Report: Interactive Job Detail Section

Noninteractive Job Detail –Sample

7-70 Performance Tools V4R2

| Job Interval Report ð9/15/98 1ð:51:36
| Non-Interactive Job Detail Page 7
| Sample Job Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| Itv Job User Job Elapsed CPU --- Nbr I/O /Sec -- -- CPU / I/O -- --- Printer ---
| End Name Name Number Pool Type Pty Time Util Sync Async Lgl Sync Async Lines Pages
| ----- ---------- ---------- ------ ---- ---- --- ------- ----- ----- ----- ----- ------- ------- --------- -------
| ðð:31 GCDJOB DHQB ð26331 2 B 99 15:ð2 .3ð ð ð ð ð ð ð ð
| ðð:31 LNEMONITOR C4STRESS ð26358 2 B ð5 15:ð2 .ðð ð ð ð 7 46 ð ð
| ðð:31 PPTJOB DHQB ð2633ð 2 B 99 15:ð2 .3ð ð ð ð ð ð ð ð
| ðð:31 QECS QSVSM ð26277 2 B 5ð 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 QNETWARE QSYS ð26287 2 B 5ð 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 QPFRMON QPGMR ð267ð6 2 B ðð 15:ð2 .ðð ð ð ð 5 47 ð ð
| ðð:31 QTFTPðð289 QTCP ð26413 2 B 25 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 QTFTPðð315 QTCP ð26414 2 B 25 15:ð2 .ðð ð ð ð 1ð3 1ð3 ð ð
| ðð:31 QTFTPðð324 QTCP ð26422 2 B 25 15:ð2 .ðð ð ð ð 56 113 ð ð
| ðð:31 QTFTPð1233 QTCP ð26433 2 B 25 15:ð2 .ðð ð ð ð 113 113 ð ð
| ðð:31 QTMSNMP QTCP ð264ð5 2 B 35 15:ð2 8.3ð ð ð ð ð ð ð ð
| ðð:31 QTMSNMPRCV QTCP ð264ð9 2 B 5ð 15:ð2 1.4ð ð ð ð ð ð ð ð
| ðð:31 QTSMTPSRVR QTCP ð264ð8 2 B 35 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 Q1PSCH QPM4ðð ð26284 2 A 5ð 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 SHNEPMð1 DHQB ð26326 2 B 5ð 15:ð2 13.ðð 3 ð ð 36 187 ð ð
| ðð:31 SHNEPMð2 DHQB ð26327 2 B 5ð 15:ð2 .ðð ð ð ð ð ð ð ð
| ðð:31 SNMP_DAD DHQB ð26697 2 B 5ð 15:ð2 .5ð ð ð ð 697 2,442 163 3
| ðð:31 SNMP_R5C DHQB ð26694 2 B 5ð 15:ð2 .ðð ð ð ð ð ð 26 ð
| ðð:31 SNMP_R7D DHQB ð26435 2 B 5ð 15:ð2 .5ð ð ð ð 1,538 4,616 163 2
| ðð:31 SNMP_R7D DHQB ð26456 2 B 5ð 15:ð2 .4ð ð ð ð 556 4,454 157 3
| ðð:31 SNMP_R7D DHQB ð26473 2 B 5ð 15:ð2 .5ð ð ð ð 2,413 4,826 163 3
| ðð:31 SNMP_R7D DHQB ð26493 2 B 5ð 15:ð2 .5ð ð ð ð 1,527 2,291 163 3
| ðð:31 SNMP_R7D DHQB ð26511 2 B 5ð 15:ð2 .5ð ð ð ð 4,534 4,534 162 3
| ðð:31 SNMP_R7D DHQB ð26535 2 B 5ð 15:ð2 .4ð ð ð ð 249 4,485 162 3
| ðð:31 SNMP_R7D DHQB ð26555 2 B 5ð 15:ð2 .5ð ð ð ð 1,621 4,863 164 3
| Itv End -- Interval end time (hour and minute)
| Job Name -- Job name
| User Name -- User name
| Job Number -- Job number
| Pool -- Pool in which the job ran
| Type -- Type and subtype of the job
| Pty -- Priority of the job
| Elapsed Time -- Elapsed time for job during interval (minutes and seconds)
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| Sync I/O /Sec -- Average number of synchronous disk I/O operations per second
| Async I/O /Sec -- Average number of asynchronous disk I/O operations per second
| Lgl I/O /Sec -- Average number of logical disk I/O operations per second
| CPU/ Sync I/O -- Avg number of CPU milliseconds per synchronous disk I/O operation
| CPU/ Async I/O -- Avg number of CPU milliseconds per asynchronous disk I/O operation
| Printer Lines -- Number of lines printed
| Printer Pages -- Number of pages printed

Figure 7-56. Job Interval Report: Noninteractive Job Detail Section

Job Interval Report Selection Criteria: Select Parameters–Sample

Chapter 7. Performance Reports—Manager Feature 7-71

 Job Interval Report 11/19/95 13:28:36
Report Selection Criteria Page 3
User-Selected Report Title

Member . . . : R2162B Model/Serial . : 53ð-ðð1ð/1ð-16B7D Main storage . . : 4ð96.ð M Started . . . : ð3/21/96 ð8:37:14
Library . . : R21626ð827 System name . . : ABSYSTEM Version/Release . : 3/7.ð Stopped . . . : ð3/21/96 ð9:ð7:15

Select Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Select parameters were chosen.

Figure 7-57. Job Interval Report: Select Parameters

Job Interval Report Selection Criteria: Omit Parameters–Sample

Omit Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

nnnnnn User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Omit parameters were chosen.

Figure 7-58. Job Interval Report: Omit Parameters

Pool Interval Report

7-72 Performance Tools V4R2

Printing the Pool Interval Report
Use the Print Pool Report (PRTPOLRPT) command, or select option 6 (Pool report)
on the Print Performance Reports display.

What Is the Pool Interval Report?

The Pool Report contains a section on subsystem activity and a section on pool
activity. Data is shown for each sample interval. Because the report can be long,
you may want to limit the output by selecting the intervals and jobs you want to
include.

If a value is too large to fit in the allotted space, a 9 is printed in each numeric field
in the report.
Note: The Pool Report is produced from the sample data collected by the
STRPFRMON command.

For samples of each section of the Pool Report, see “Sample Pool Interval
Reports.”

For definitions of specific columns in the reports, see “Performance Report

Columns” on page 7-102.

Subsystem Activity
The Subsystem Activity section of the Pool Interval Report gives the performance
information on the subsystems during each selected interval. One line is printed for
each subsystem and active pool combination that existed during each selected
interval.

See the sample report shown in Figure 7-59 on page 7-74.

Pool Activity
The Pool Activity section of the Pool Interval Report gives the performance informa-
tion on the storage pools at various time intervals. One line is printed for each
active pool that existed during each selected interval.

See the sample report shown in Figure 7-60 on page 7-75.

Report Selection Criteria

The Report Selection Criteria section of the Pool Interval Report gives the selection
values you chose to produce the report.

See the sample report shown in Figure 7-61 on page 7-76.

Sample Pool Interval Reports

See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Chapter 7. Performance Reports—Manager Feature 7-73

 Subsystem Activity–Sample
| Pool Interval Report ð9/15/98 1ð:52:24
| Subsystem Activity Page 1
| Sample Pool Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| --------- Physical I/O per Transaction -------- ---------------- Job Maximums ----------------
| Itv Subsystem CPU ----- Synchronous ----- ---- Asynchronous ----- CPU Phy
| End Name PL Util Tns DBR DBW NDBR NDBW DBR DBW NDBR NDBW Util I/O Tns Rsp A-W W-I A-I
| ----- ---------- -- ---- ---------- ----- ----- ----- ----- ----- ----- ----- ----- ---- ---------- ----- ------ ------ ---- ----
| ðð:31 QBATCH 2 22.ð ð 13.ð 3,877 ð .ðð 328 ð ð
| ðð:31 QSOC 2 .ð ð .ð ð ð .ðð 88 ð ð
| ðð:31 QSYSWRK 2 9.8 ð 8.3 6 ð .ðð 49,835 ð ð
| ðð:31 Q1PGSCH 2 .ð ð .ð 4ð ð .ðð 15 ð ð
| ðð:46 QBATCH 2 21.7 ð 12.5 5,182 ð .ðð 327 ð ð
| ðð:46 QSOC 2 .ð ð .ð ð ð .ðð 88 ð ð
| ðð:46 QSYSWRK 2 1ð.ð ð 8.4 339 ð .ðð 49,954 ð ð
| ðð:46 Q1PGSCH 2 .ð ð .ð 17ð ð .ðð 15 ð ð
| ð1:ð1 QBATCH 2 21.8 ð 12.4 4,ð88 ð .ðð 327 ð ð
| ð1:ð1 QSOC 2 .ð ð .ð ð ð .ðð 87 ð ð
| ð1:ð1 QSYSWRK 2 9.8 ð 8.2 93 ð .ðð 49,344 ð ð
| ð1:ð1 Q1PGSCH 2 .ð ð 8.8 164 ð .ðð 17 ð ð
| ð1:16 QBATCH 2 19.6 ð 1ð.8 4,864 ð .ðð 322 ð ð
| ð1:16 QSOC 2 .ð ð .ð ð ð .ðð 88 ð ð
| ð1:16 QSYSWRK 2 11.5 ð 9.7 4 ð .ðð 59,578 ð ð
| ð1:16 Q1PGSCH 2 .ð ð .ð 154 ð .ðð 15 ð ð
| ð1:31 QBATCH 2 2ð.6 ð 1ð.2 4,188 ð .ðð 323 ð ð
| ð1:31 QSOC 2 .ð ð .ð ð ð .ðð 88 ð ð
| ð1:31 QSYSWRK 2 1ð.8 ð 9.2 4 ð .ðð 54,293 ð ð
| Itv End -- Interval end time (hour and minute)
| Subsystem Name -- Subsystem name
| PL -- Pool in which the jobs in the subsystem ran
| CPU Util -- Average CPU utilization by the transactions in the subsystem. This is the average of all processors
| Tns -- Number of transactions in the subsystem
| Physical I/O per Trans -- Average physical disk I/O operations per transaction
| Synchronous DBR -- Average synchronous data base reads per transaction
| Synchronous DBW -- Average synchronous data base writes per transaction
| Synchronous NDBR -- Average synchronous non-data base reads per transaction
| Synchronous NDBW -- Average synchronous non-data base writes per transaction
| Asynchronous DBR -- Average asynchronous data base reads per transaction
| Asynchronous DBW -- Average asynchronous data base writes per transaction
| Asynchronous NDBR -- Average asynchronous non-data base reads per transaction
| Asynchronous NDBW -- Average asynchronous non-data base writes per transaction
| Job Maximums -- Maximum values by a job in the subsystem
| CPU Util -- Highest percentage CPU utilization
| Phy I/O -- Most physical disk I/O requests
| Tns -- Most transactions
| Rsp -- Highest average response time (seconds)
| A-W -- Most active-to-wait transitions
| W-I -- Most wait-to-ineligible transitions
| A-I -- Most active-to-ineligible transitions

Figure 7-59. Pool Interval Report: Subsystem Activity

Pool Activity–Sample

7-74 Performance Tools V4R2

| Pool Interval Report ð9/15/98 1ð:52:24
| Pool Activity Page 12
| Sample Pool Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| -------- Physical I/O per Transaction --------- ---------------- Job Maximums --------------
| Itv Act Size CPU ----- Synchronous ----- ----- Asynchronous ---- CPU Phy
| End PL Lvl (K) Util Tns DBR DBW NDBR NDBW DBR DBW NDBR NDBW Util I/O Tns Rsp A-W W-I A-I
| ----- -- ----- --------- ---- -------- ----- ----- ----- ----- ----- ----- ----- ----- ---- --------- ----- ------ ------ ---- ----
| ðð:31 2 44 1752ð4 31.8 ð 13.ð 3,877 ð .ðð 49,835 ð ð
| ðð:46 2 44 175976 31.9 ð 12.5 5,182 ð .ðð 49,954 ð ð
| ð1:ð1 2 44 177768 31.7 ð 12.4 4,ð88 ð .ðð 49,344 ð ð
| ð1:16 2 44 1785ðð 31.2 ð 1ð.8 4,864 ð .ðð 59,578 ð ð
| ð1:31 2 44 175364 31.5 ð 1ð.2 4,188 ð .ðð 54,293 ð ð
| ð1:46 2 44 1769ð8 31.5 ð 8.8 1,952 ð .ðð 48,791 ð ð
| ð2:ð1 2 44 176864 31.6 ð 8.8 1,355 ð .ðð 47,491 ð ð
| ð2:16 2 44 1776ð8 32.ð ð 11.2 2,223 ð .ðð 42,851 ð ð
| ð2:31 2 44 176896 31.9 ð 11.7 3,1ð9 ð .ðð 45,664 ð ð
| ð2:46 2 44 175756 32.ð ð 12.3 1,487 ð .ðð 47,998 ð ð
| ð3:ð1 2 44 175296 31.7 ð 12.2 5,628 ð .ðð 5ð,733 ð ð
| Itv End -- Interval end time (hour and minute)
| PL -- Pool identifier
| Act Lvl -- Activity level of the pool
| Size (K) -- Size of the pool (kilobytes)
| CPU Util -- Average CPU utilization by the transactions in the pool. This is the average of all processors
| Tns -- Number of transactions in the pool
| Physical I/O per Trans -- Average physical disk I/O operations per transaction in the pool
| Synchronous DBR -- Average synchronous data base reads per transaction
| Synchronous DBW -- Average synchronous data base writes per transaction
| Synchronous NDBR -- Average synchronous non-data base reads per transaction
| Synchronous NDBW -- Average synchronous non-data base writes per transaction
| Asynchronous DBR -- Average asynchronous data base reads per transaction
| Asynchronous DBW -- Average asynchronous data base writes per transaction
| Asynchronous NDBR -- Average asynchronous non-data base reads per transaction
| Asynchronous NDBW -- Average asynchronous non-data base writes per transaction
| Job Maximums -- Maximum values by a job in the pool
| CPU Util -- Highest percentage CPU utilization
| Phy I/O -- Most physical disk I/O requests
| Tns -- Most transactions
| Rsp -- Highest average response time (seconds)
| A-W -- Most active-to-wait transitions
| W-I -- Most wait-to-ineligible transitions
| A-I -- Most active-to-ineligible transitions

Figure 7-60. Pool Interval Report: Pool Activity

Report Selection Criteria–Sample

Chapter 7. Performance Reports—Manager Feature 7-75

 Pool Interval Report ð5/ð7/96 13:43:2ð
Report Selection Criteria Page 3
User-Selected Report Title

Member . . . : R2162B Model/Serial . : 53ð-ðð1ð/1ð-16B7D Main storage . . : 4ð96.ð M Started . . . : ð3/21/96 ð8:37:14
Library . . : R21626ð827 System name . . : ABSYSTEM Version/Release . : 3/7.ð Stopped . . . : ð3/21/96 ð9:ð7:15

Select Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

User7 User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Select parameters were chosen.

Omit Parameters

Pools - ð1 ð2 ð3 ð4 ð5 ð6 ð7 ð8 ð9 1ð 11 12 13 14 15 16

Jobs - ð12345/Useridwxyz/Jobname123
987654/Useridabcd/Jobname456

User IDs - User1 User2 User3 User4 User5 User6

nnnnnn User8 User9 User1ð User11 User12

Subsystems - Subsystem1 Subsystem2 Subsystem3 Subsystem4 Subsystem5 Subsystem6

Subsystem7 Subsystem8 Subsystem9 Subsystema Subsystemb Subsystemc

Communications Lines - Line1 Line2 Line3 Line4 Line5 Line6

Line7 Line8 Line9 Line1ð Line11 Line12

Control Units - Ctlr1 Ctlr2 Ctlr3 Ctlr4 Ctlr5 Ctlr6

Ctlr7 Ctlr8 Ctlr9 Ctlr1ð Ctlr11 Ctlr12

Functional Areas - Accounting Payroll Research

Development ProjectX MrNolansStaff

- No Omit parameters were chosen.

Figure 7-61. Pool Interval Report: Report Selection Criteria

Resource Interval Report

Printing the Resource Interval Report

Use the Print Resource Report (PRTRSCRPT) command, or select option 7
(Resource report) on the Print Performance Reports display.

What Is the Resource Interval Report?

The six major sections of the Resource Interval Report provide resource information
on all or selected intervals. Because the report can be long, you may want to limit
the output by selecting the intervals you want to include.

If a value is too large to fit in the allotted space, a 9 is printed in each numeric field
in the report.

7-76 Performance Tools V4R2

 Sample Data: this report is produced from the sample data collected with the
STRPFRMON command.

Disk Utilization Summary

The Disk Utilization Summary section of the Resource Interval Report gives
detailed disk information by time intervals.

Information is shown for all disk arms that are configured on the system. Also, the
disk arm with the highest utilization and the disk arm with the highest average seek
time for each time interval are shown. Consistent disk arm utilization at or above
the threshold value will affect system performance and cause longer response
times and/or less throughput.

See the sample report shown in Figure 7-62 on page 7-80.

Disk Utilization Detail

The Disk Utilization Detail section of the Resource Interval Report gives detailed
disk information for the selected time intervals.

Information is shown for each disk arm that is configured on the system. Con-
sistent disk arm utilization at or above the threshold value will affect system per-
formance and cause longer response times and/or less throughput.

See the sample report shown in Figure 7-63 on page 7-81.

Communications Line Detail

A Communications Line Detail section of a Resource Interval Report contains infor-
mation about the line activity when performance data was collected for the specified
member. One detail section is produced for each protocol in use on the lines for
which data was collected. Figure 7-64 on page 7-82 through Figure 7-70 on
page 7-85 are samples of the detail sections for the communications protocols.
Note: Each section appears only if you have communications lines using that par-
ticular protocol.

SDLC Protocol
Figure 7-64 on page 7-82 is a sample of the report section for communications
lines using the synchronous data link control (SDLC) protocol. The data in this
example is sorted by the data collection interval end times.

X.25 Protocol
Figure 7-65 on page 7-82 is a sample of the report section for communications
lines using the X.25 protocol.

TRLAN Protocol
Figure 7-66 on page 7-83 is a sample of the report section for communications
lines using the token-ring local area network (TRLAN) protocol.

Chapter 7. Performance Reports—Manager Feature 7-77

 ELAN Protocol
Figure 7-67 on page 7-83 is a sample of the report section for communications
lines using the Ethernet local area network (ELAN) protocol.

DDI Protocol
Figure 7-68 on page 7-84 is a sample of the report section for communications
lines using the distributed data interface (DDI) protocol.

FRLY Protocol
Figure 7-69 on page 7-84 is a sample of the report section for communications
lines using the frame relay (FRLY) protocol.

ASYNC Protocol
Figure 7-70 on page 7-85 is a sample of the report section for communications
lines using the asynchronous (ASYNC) protocol.
Note: A protocol data unit (PDU) for asynchronous communications is a variable-
length unit of data that is ended by a protocol control character or by the
size of the buffer.

BSC Protocol
Figure 7-71 on page 7-85 is a sample of the report section for communications
lines using the binary synchronous communications (BSC) protocol.

ISDN Network Interface

Figure 7-72 on page 7-86 is a sample of the report section for the integrated ser-
vices digital network (ISDN) network interface.

Network Interface Maintenance Channel for ISDN

Figure 7-73 on page 7-86 is a sample of the report section for the network inter-
face maintenance channel for the ISDN protocol.

IDLC Protocol
Figure 7-74 on page 7-87 and Figure 7-75 on page 7-87 are samples of the report
section for communications lines using the ISDN data link control (IDLC) protocol.
Figure 7-75 on page 7-87 indicates which B-channel the IDLC line was using
during the interval.

Related Information

Information regarding ISDN can be found in the ISDN Support book.

IOP Utilizations
The IOP Utilizations section of the Resource Interval Report contains a combination
of the following input/output processor (IOP) utilizations:
Ÿ Disk IOP utilizations
Gives input/output processor (IOP) utilization for direct access storage devices
(DASDs). Consistent Disk IOP utilization at or above the threshold value affects
system performance and causes longer response times and/or less throughput.
Ÿ Multifunction IOP utilizations

7-78 Performance Tools V4R2

 Gives input/output processor (IOP) utilization for DASD, communication, and
local workstation devices. Consistent utilization at or above the threshold value
affects system performance and causes longer response times and/or less
throughput.
Ÿ Communications IOP utilizations
Gives communications input/output processor (IOP) utilization.
Ÿ Local work station IOP utilizations
Gives input/output processor (IOP) utilization for local workstation devices.

See the sample report shown in shown in Figure 7-76 on page 7-88.
Note: The total for the I/O processor utilization oftentimes does not match the sum
of the three columns (IOP Processor Util Comm, IOP Processor Util LWSC,
and IOP Processor Util DASD). This mismatch is caused by the utilization of
other small components, such as system time.

Local Work Station Response Times

The local work station response times section provides the following for each data
collection interval:
Ÿ Local work station IOP utilization
Ÿ Number of work stations active on each controller
Ÿ Range of response times for the work stations
Ÿ Average response time for the work stations

The response time intervals are set up on the system STRPFRMON command.
The values for the response times may vary depending on the values you use in
the command.

See the sample report shown in Figure 7-77 on page 7-89.

Remote Work Station Response Times

The remote work station response times section gives the following for each data
collection interval:
Ÿ Number of work stations active on each controller
Ÿ Range of response times for the work stations
Ÿ Average response time for the work stations

The response time intervals are set up on the system STRPFRMON command.
The values for the response times may vary depending on the values you use in
the command.
Note: This section appears only if a 5494 remote controller is included in the data
collection.

See the sample report shown in Figure 7-78 on page 7-89.

Chapter 7. Performance Reports—Manager Feature 7-79

 Sample Resource Interval Reports
See “Performance Report Columns” on page 7-102 for an alphabetized list con-
taining definitions for each column in the reports.

Disk Utilization Summary–Sample

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Disk Utilization Summary Page 1
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| Average Average Average High High High Disk
| Itv Average Reads Writes K Per Avg High Util Srv Srv Space
| End I/O /Sec /Sec /Sec I/O Util Util Unit Time Unit Used
| ----- --------- -------- -------- ------- ---- ---- ---- ----- ---- -----------
| 13:14 4ð.7 18.7 21.9 6.1 16.5 24.4 ððð2 .ð138 ððð2 9,682
| 13:19 32.7 16.3 16.3 5.6 13.2 2ð.3 ððð2 .ð153 ððð2 9,691
| 13:24 37.6 19.4 18.1 5.8 16.ð 22.7 ððð2 .ð143 ððð2 9,697
| 13:29 35.7 18.2 17.5 6.1 14.8 19.5 ððð2 .ð127 ððð2 9,7ð9
| 13:34 25.9 12.7 13.2 5.5 11.3 17.1 ððð2 .ð152 ððð2 9,718
| 13:38 38.3 17.8 2ð.4 6.1 16.9 24.7 ððð2 .ð158 ððð2 9,721
| --------- -------- -------- ------- ----
| Average: 35.1 17.2 17.9 5.9 14.8
| Itv End -- Interval end time (hour and minute)
| Average Phys I/O /Sec -- Average number of physical I/O operations per second
| Average Reads / Sec -- Average number of physical reads per second
| Average Writes /Sec -- Average number of physical writes per second
| Average K Per I/O -- Average number of kilobytes (1ð24) per I/O operation
| Avg Util -- Average percent utilization of all disk arms
| High Util -- Highest percent utilization for a disk arm
| High Util Unit -- Disk arm with the highest utilization percent
| High Srv Time -- Highest average service time in seconds
| High Srv Unit -- Disk arm with the highest service time
| Disk Space Used -- Total disk space used in millions of bytes

Figure 7-62. Resource Interval Report: Disk Utilization Summary

Disk Utilization Detail–Sample

7-80 Performance Tools V4R2

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Disk Utilization Detail Page 2
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| IOP Name/ ASP Itv -------- I/O Per Second --------- K Per Dsk CPU Queue Avg Time Per I/O
| Unit (Model) Id End Total Reads Writes I/O Util Util Length Service Wait
| ---- ----------------- --- ----- ---------- --------- --------- ------- ------- ---- ------- ------- ----
| ððð1 CMBð1 (66ð7) ð1 13:14 13.899 7.18ð 6.719 6.8 .ð 15.7 .24 .ð112 .ðð71
| 13:19 1ð.797 5.72ð 5.ð76 5.9 .ð 1ð.3 .15 .ðð95 .ðð53
| 13:24 12.61ð 6.943 5.666 6.2 .ð 14.8 .2ð .ð117 .ðð48
| 13:29 12.646 6.763 5.883 6.7 .ð 15.2 .26 .ð12ð .ðð97
| 13:34 8.744 5.5ð8 3.235 5.9 .ð 1ð.8 .15 .ð123 .ðð63
| 13:38 13.557 6.4ð9 7.147 6.1 .ð 15.7 .29 .ð115 .ð1ð6
| ---------- --------- --------- ------- ---- ------- ------- ----
| Unit Average: 12.ð13 6.419 5.594 6.3 13.7 .22 .ð113 .ðð73
| ððð2 CMBð1 (66ð7) ð1 13:14 17.685 6.719 1ð.966 6.ð .ð 24.4 .49 .ð137 .ð148
| 13:19 13.262 6.ððð 7.262 5.8 .ð 2ð.3 .32 .ð153 .ðð94
| 13:24 15.796 7.236 8.56ð 5.8 .ð 22.7 .37 .ð143 .ðð97
| 13:29 15.346 6.69ð 8.656 6.1 .ð 19.5 .31 .ð127 .ðð84
| 13:34 11.255 4.196 7.ð59 5.4 .ð 17.1 .25 .ð151 .ðð83
| 13:38 15.6ð5 6.653 8.952 6.3 .ð 24.7 .47 .ð158 .ð153
| ---------- --------- --------- ------- ---- ------- ------- ----
| Unit Average: 14.8ð7 6.24ð 8.567 5.9 21.4 .37 .ð144 .ð1ð9
| ððð3 CMBð1 (66ð6) ð1 13:14 9.13ð 4.882 4.247 5.3 .ð 9.5 .11 .ð1ð4 .ðð27
| 13:19 8.684 4.647 4.ð36 5.1 .ð 9.1 .11 .ð1ð4 .ðð35
| 13:24 9.223 5.3ðð 3.923 5.2 .ð 1ð.6 .15 .ð114 .ðð6ð
| 13:29 7.776 4.773 3.ðð3 5.4 .ð 9.8 .14 .ð126 .ðð61
| 13:34 5.96ð 3.ðð9 2.95ð 5.ð .ð 6.1 .ð8 .ð1ð2 .ðð46
| 13:38 9.177 4.819 4.357 5.6 .ð 1ð.3 .12 .ð112 .ðð32
| ---------- --------- --------- ------- ---- ------- ------- ----
| Unit Average: 8.3ð9 4.567 3.742 5.3 9.2 .12 .ð11ð .ðð43
| Unit -- Disk arm identifier
| IOP Name/ -- Input/Output processor resource name and
| (Model) model number of the attached device
| ASP ID -- Auxiliary storage pool number
| Itv End -- Interval end time (hour and minute)
| I/O /Sec -- Average number of I/O operations per second
| Reads Per Second -- Average number of reads per second
| Writes Per Sec -- Average number of writes per second
| K Per I/O -- Average number of kilobytes (1ð24) per I/O operation
| Dsk CPU Util -- Percentage of Disk CPU Utilization
| Util -- Average percent of time disk was used (busy)
| Queue Length -- Average length of waiting queue
| Average Service Time -- Average disk service time per I/O operation
| Average Wait Time -- Average disk wait time per I/O operation

Figure 7-63. Resource Interval Report: Disk Utilization Detail

Communications Line Detail–SDLC Sample

Chapter 7. Performance Reports—Manager Feature 7-81

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 3
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = SDLC (SORT BY INTERVAL)
| Percent Percent Pct -- Congestion ---
| IOP Bytes Total I Frames Bytes Total Frames Poll Local Remote
| Itv Name/ Line Line Trnsmitd I Frames Trnsmitd Recd Frames Received Retry Not Not
| End Line Speed Util Per Sec Trnsmitd in Error Per Sec Recd in Error Time Ready Ready
| ----- ---------- -------- ------ ---------- ---------- ---------- ---------- -------- ---------- ------- ------- --------
| CCð9
| (26ð9)
| 13:14 PMSD1 19.2 4.6 49 322 ð 62 2,9ð9 ð ð ð ð
| 13:19 PMSD1 19.2 4.4 47 3ð1 ð 6ð 2,943 ð ð ð ð
| 13:24 PMSD1 19.2 5.4 56 399 ð 73 2,889 ð ð ð ð
| 13:29 PMSD1 19.2 4.ð 52 159 ð 45 3,ð29 ð ð ð ð
| 13:34 PMSD1 19.2 4.1 54 131 ð 43 3,ð74 ð ð ð ð
| 13:38 PMSD1 19.2 5.9 81 2ð6 ð 61 2,762 ð ð ð ð
| CC13
| (26ð9)
| 13:14 PMSD2 19.2 4.6 63 16ð ð 49 3,ð44 ð ð ð ð
| 13:19 PMSD2 19.2 4.4 6ð 151 ð 47 3,ð72 ð ð ð ð
| 13:24 PMSD2 19.2 5.4 73 2ðð ð 56 3,ð55 ð ð ð ð
| 13:29 PMSD2 19.2 4.ð 45 226 ð 52 2,971 ð ð ð ð
| 13:34 PMSD2 19.2 4.1 43 263 ð 55 2,966 ð ð ð ð
| 13:38 PMSD2 19.2 5.9 61 411 ð 8ð 2,587 ð ð ð ð

Figure 7-64. Resource Interval Report: Communications Line Detail - SDLC

Communications Line Detail–X.25 Sample

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 5
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = X.25 (SORT BY INTERVAL)
| Transmit/ Percent Percent
| IOP Receive/ Bytes Total I Frames Bytes Total Frames ------Reset------
| Itv Name/ Line Average Trnsmitd I Frames Trnsmitd Recd Frames Recd -----Packets-----
| End Line Speed Line Util Per Sec Trnsmitd In Error Per Sec Recd In Err Trnsmitd Recd
| ----- ---------- -------- ----------- ---------- ---------- ---------- --------- -------- -------- -------- -------
| CC13
| (26ð9)
| 13:14 PMX21 19.2 ð2/ð2/ð2 61 535 ð 52 1,ð7ð ð ð ð
| 13:14 PMX22 19.2 ð1/ð2/ð2 44 535 ð 68 1,ð7ð ð ð ð
| 13:19 PMX21 19.2 ð2/ð2/ð2 57 5ð4 ð 48 1,ðð8 ð ð ð
| 13:19 PMX22 19.2 ð1/ð2/ð2 41 5ð4 ð 63 1,ðð8 ð ð ð
| 13:24 PMX21 19.2 ð2/ð2/ð2 64 564 ð 54 1,128 ð ð ð
| 13:24 PMX22 19.2 ð1/ð2/ð2 47 564 ð 71 1,128 ð ð ð
| 13:29 PMX21 19.2 ð1/ð2/ð1 32 391 ð 49 782 ð ð ð
| 13:29 PMX22 19.2 ð1/ð1/ð1 44 391 ð 37 782 ð ð ð
| 13:34 PMX21 19.2 ð1/ð2/ð2 38 467 ð 58 934 ð ð ð
| 13:34 PMX22 19.2 ð2/ð1/ð2 52 467 ð 44 934 ð ð ð
| 13:38 PMX21 19.2 ð2/ð4/ð3 69 751 ð 1ð6 1,5ð2 ð ð ð
| 13:38 PMX22 19.2 ð3/ð3/ð3 95 751 ð 8ð 1,5ð2 ð ð ð

Figure 7-65. Resource Interval Report: Communications Line Detail - X.25

Communications Line Detail–TRLAN Sample

7-82 Performance Tools V4R2

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 7
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = TRLAN (SORT BY INTERVAL)
| ------- Congestion -------
| IOP I Frames I Frames -- Local -- -- Remote -- Rsp Remote LAN
| Itv Name/ Line Line Trnsmitd Recd Not Seq Not Seq Frame Timer -- Pct Frames -- MAC
| End Line Speed Util Per Sec Per Sec Ready Error Ready Error Retry Ended Trnsmitd Recd Errors
| ----- ---------- -------- ------- ---------- ---------- ----- ----- ----- ----- ------- ------ -------- ---- -------
| CCð4
| (2619)
| 13:14 .ð .ð ð ð ð ð ð ð ð ð ð 97 ð
| 13:19 .ð .ð ð ð ð ð ð ð ð ð ð 98 ð
| 13:24 .ð .ð ð ð ð ð ð ð ð ð ð 98 ð
| 13:29 .ð .ð ð ð ð ð ð ð ð ð ð 98 ð
| 13:34 .ð .ð ð ð ð ð ð ð ð ð ð 98 ð
| 13:38 .ð .ð ð ð ð ð ð ð ð ð ð 97 ð
| CCð6
| (2626)
| 13:14 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 3 8ð ð
| 13:19 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 6 85 ð
| 13:24 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 5 85 ð
| 13:29 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 4 83 3
| 13:34 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 6 86 ð
| 13:38 TRNLINE 16ððð.ð .ð ð ð ð ð ð ð ð ð 6 86 ð
| CCð7
| (2619)
| 13:14 PMEC1 16ððð.ð .ð 3 3 ð ð ð ð ð ð 1ðð 1ðð ð
| 13:19 PMEC1 16ððð.ð .ð 3 2 ð ð ð ð ð ð 1ðð 1ðð ð
| 13:24 PMEC1 16ððð.ð .ð 1 1 ð ð ð ð ð ð 1ðð 1ðð ð

Figure 7-66. Resource Interval Report: Communications Line Detail - TRLAN

Communications Line Detail–ELAN Sample

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 1ð
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = ELAN (SORT BY INTERVAL)
| -------------- Congestion --------------
| IOP I Frames I Frames ----- Local ----- ----- Remote ----- Rsp
| Itv Name/ Line Line Trnsmitd Recd Not Seq Not Seq Frame Timer
| End Line Speed Util Per Sec Per Sec Ready Error Ready Error Retry Ended
| ----- ---------- -------- ------ ---------- ---------- ------- ------- ------- ------- ------- -------
| CCð3
| (2617)
| 13:14 PMET2 1ðððð.ð .ð 3 3 ð ð ð ð ð ð
| 13:19 PMET2 1ðððð.ð .ð 2 2 ð ð ð ð ð ð
| 13:24 PMET2 1ðððð.ð .ð 2 1 ð ð ð ð ð ð
| 13:29 PMET2 1ðððð.ð .ð 2 2 ð ð ð ð ð ð
| 13:34 PMET2 1ðððð.ð .ð 1 1 ð ð ð ð ð ð
| 13:38 PMET2 1ðððð.ð .ð ð ð ð ð ð ð ð ð
| CCð5
| (2617)
| 13:14 PMET1 1ðððð.ð .ð 3 3 ð ð ð ð ð ð
| 13:19 PMET1 1ðððð.ð .ð 2 2 ð ð ð ð ð ð
| 13:24 PMET1 1ðððð.ð .ð 1 2 ð ð ð ð ð ð
| 13:29 PMET1 1ðððð.ð .ð 2 2 ð ð ð ð ð ð
| 13:34 PMET1 1ðððð.ð .ð 1 1 ð ð ð ð ð ð
| 13:38 PMET1 1ðððð.ð .ð ð ð ð ð ð ð ð ð

Figure 7-67. Resource Interval Report: Communications Line Detail–ELAN

Chapter 7. Performance Reports—Manager Feature 7-83

 Communications Line Detail–DDI Sample
| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 12
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = DDI (SORT BY INTERVAL)
| ------- Congestion -------
| IOP I Frames I Frames -- Local -- -- Remote -- Rsp
| Itv Name/ Line Line Trnsmitd Recd Not Seq Not Seq Frame Timer MAC
| End Line Speed Util Per Sec Per Sec Ready Error Ready Error Retry Ended Errors
| ----- ---------- -------- ------ ---------- ---------- ----- ----- ----- ----- ------- ------ -------
| CCð1
| (2618)
| 13:14 PMDD1 1ððððð.ð .ð 3 3 ð ð ð ð ð ð ð
| 13:19 PMDD1 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð
| 13:24 PMDD1 1ððððð.ð .ð 2 2 ð ð ð ð ð ð ð
| 13:29 PMDD1 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð
| 13:34 PMDD1 1ððððð.ð .ð 1 1 ð ð ð ð ð ð ð
| 13:38 PMDD1 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð
| CCð2
| (2618)
| 13:14 PMDD2 1ððððð.ð .ð 3 3 ð ð ð ð ð ð ð
| 13:19 PMDD2 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð
| 13:24 PMDD2 1ððððð.ð .ð 2 2 ð ð ð ð ð ð ð
| 13:29 PMDD2 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð
| 13:34 PMDD2 1ððððð.ð .ð 1 1 ð ð ð ð ð ð ð
| 13:38 PMDD2 1ððððð.ð .ð ð ð ð ð ð ð ð ð ð

Figure 7-68. Resource Interval Report: Communications Line Detail–DDI

Communications Line Detail–FRLY Sample

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 14
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = FRLY (SORT BY INTERVAL)
| ------- Congestion -------
| IOP I Frames I Frames -- Local -- -- Remote -- Rsp
| Itv Name/ Line Line Trnsmitd Recd Not Seq Not Seq Frame Timer MAC
| End Line Speed Util Per Sec Per Sec Ready Error Ready Error Retry Ended Errors
| ----- ---------- -------- ------ ---------- ---------- ----- ----- ----- ----- ------- ------ -------
| CC1ð
| (2666)
| 13:14 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:19 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:24 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:29 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:34 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:38 PMFR1 56.ð .ð ð ð ð ð ð ð ð ð ð
| CC11
| (2666)
| 13:14 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:19 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:24 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:29 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:34 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð
| 13:38 PMFR2 56.ð .ð ð ð ð ð ð ð ð ð ð

Figure 7-69. Resource Interval Report: Communications Line Detail–FRLY

Communications Line Detail–ASYNC Sample

7-84 Performance Tools V4R2

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 16
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = ASYNC (SORT BY INTERVAL)
| IOP Bytes Bytes Total Pct PDUs
| Itv Name/ Line Line Transmitted Received PDUs Received
| End Line Speed Util Per Sec Per Sec Received in Error
| ----- ---------- -------- ------ --------------- ------------ ------------ ------------
| CCð9
| (26ð9)
| 13:14 PMAS1 1.2 17.6 26 ð 1ð6 ð
| 13:19 PMAS1 1.2 1ð.ð 14 ð 64 ð
| 13:24 PMAS1 1.2 7.5 11 ð 55 ð
| 13:29 PMAS1 1.2 13.2 19 ð 72 ð
| 13:34 PMAS1 1.2 11.8 17 ð 47 ð
| 13:38 PMAS1 1.2 7.8 11 ð 36 ð
| CC13
| (26ð9)
| 13:14 PMAS2 1.2 17.7 ð 26 79 ð
| 13:19 PMAS2 1.2 1ð.2 ð 15 47 ð
| 13:24 PMAS2 1.2 7.5 ð 11 32 ð
| 13:29 PMAS2 1.2 13.2 ð 19 57 ð
| 13:34 PMAS2 1.2 11.8 ð 17 54 1
| 13:38 PMAS2 1.2 7.8 ð 11 29 ð

Figure 7-70. Resource Interval Report: Communications Line Detail–ASYNC

Communications Line Detail–BSC Sample

| Resource Interval Report ð9/18/98 14:ð6:ðð
| Communications Line Detail Page 18
| Sample Resource Interval Report
| Member . . . : PMISTGA1 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . : ð8/11/98 13:ð9:ð4
| Library . . : PM42CRT System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/11/98 13:38:4ð
| PROTOCOL = BSC (SORT BY INTERVAL)
| Pct Data Pct Data
| IOP Bytes Total Data Characters Bytes Total Data Characters
| Itv Name/ Line Line Transmitted Characters Transmitted Received Characters Received Line
| End Line Speed Util Per Sec Transmitted in Error Per Sec Received in Error Errors
| ----- ---------- -------- ------ ------------- ------------- ------------- ---------- ------------ ------------ --------
| CC13
| (26ð9)
| 13:14 PMBS1 19.2 .9 7 2,36ð ð 13 4,124 ð ð
| 13:14 PMBS2 19.2 .9 13 4,124 ð 7 2,36ð ð ð
| 13:19 PMBS1 19.2 1.1 9 2,99ð ð 17 5,226 ð ð
| 13:19 PMBS2 19.2 1.1 17 5,226 ð 9 2,99ð ð ð
| 13:24 PMBS1 19.2 .9 8 2,568 ð 15 4,488 ð ð
| 13:24 PMBS2 19.2 .9 15 4,488 ð 8 2,568 ð ð
| 13:29 PMBS1 19.2 1.1 1ð 3,1ð3 ð 18 5,423 ð ð
| 13:29 PMBS2 19.2 1.1 18 5,423 ð 1ð 3,1ð3 ð ð
| 13:34 PMBS1 19.2 1.2 11 3,424 ð 19 5,984 ð ð
| 13:34 PMBS2 19.2 1.2 19 5,984 ð 11 3,424 ð ð
| 13:38 PMBS1 19.2 1.ð 9 2,463 ð 15 4,3ð2 ð ð
| 13:38 PMBS2 19.2 1.ð 15 4,3ð2 ð 9 2,463 ð ð

Figure 7-71. Resource Interval Report: Communications Line Detail - BSC

Communications Line Detail–ISDN Network Interface Sample

Chapter 7. Performance Reports—Manager Feature 7-85

 Resource Interval Report ð9/23/98 ð6:14:ð4
Communications Line Detail Page 15
Sample Resource Interval Report
Member . . . : ISDNDATA Model/Serial . : 5ðð-2142/1ð-1ðDFD Main storage . . : 32ð.ð M Started . . . : ð8/14/98 13:3ð:23
Library . . : ISDNDATA System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð8/14/98 13:45:27
PROTOCOL = ISDN NETWORK INTERFACE (SORT BY INTERVAL)
IOP --Outgoing--- ---Incoming--- LAPD LAPD Pct LAPD LAPD Pct
Name/ ---Calls----- ----Calls----- Total Frames Total Frames Loss of Local
Itv Network Line Pct Pct Frames Trnsmitd Frames Recd Frame End Code Collision
End Interface Speed Total Retry Total Reject Trnsmitd Again Recd in Error Alignment Violation Detect
----- ---------- ----- ------ ----- ------ ------ -------- -------- ------- -------- --------- --------- ---------
CCð5
(26ð5)
13:35 X31Nðð 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
13:35 X31Nð1 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
13:4ð X31Nðð 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
13:4ð X31Nð1 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
13:45 X31Nðð 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
13:45 X31Nð1 16.3 ð ð ð ð 6ð ð 6ð ð ð ð ð
Itv End -- End time of the data collection interval or time that vary
off occurred
IOP Name/ -- IOP resource name and model number, Network interface description
Network Interface
Line Speed -- Line speed (1ððð bits per second)
Outgoing Calls Total -- Number of outgoing call attempts
Outgoing Calls -- Percent of outgoing calls that were rejected by the network
Pct Retry
Incoming Calls Total -- Number of incoming call attempts
Incoming Calls -- Percent of incoming calls that were rejected
Pct Reject
LAPD Total Frames -- Number of frames transmitted (applies to D-channel only)
Trnsmitd
LAPD Pct Frames -- Percent frames re-transmitted due to error (applies to
Trnsmitd Again D-channel only)
LAPD Total Frames -- Number of frames received (applies to D-channel only)
Recd
LAPD Pct Frames -- Percent frames received in error (applies to D-channel only)
Recd in Error
Loss of Frame -- Number of times a time period equivalent to two 48 bit frames
Alignment elapsed without detecting valid pairs of line code violations
Local End Code -- Number of unintended code violations detected by the TE
Violation for frames received on the T interface
Collision Detect -- Number of times that a transmitted frame corrupted by
another frame was detected

Figure 7-72. Resource Interval Report: Communications Line Detail - ISDN Network Interface

Communications Line Detail–NWI Maintenance Sample

Resource Interval Report 11/1ð/95 ð8:ðð:33
Communications Line Detail Page 13
User-Selected Report Title

Member . . . : MONDAY Model/Serial . : 2ðð-2ð5ð/1ð-15ðð5ðð Main storage . . : 16ð.ð M Started . . . : 11/ð2/95 14:31:23
Library . . : QPFRDATA System name . . : ABSYSTEM Version/Release . : 3/ 6.ð Stopped . . . : 11/ð2/95 16:26:12

PROTOCOL = NWI MAINTENANCE CHANNEL (SORT BY INTERVAL)

IOP Percent
Name/ Percent Severely -----Detected Access----- Far End
Itv Network Line Errored Errored ----Transmission Error--- Code
End Interface Speed Seconds Seconds In Out Violation
----- ---------- -------- ------- -------- ---------- ---------- ----------
CC11
(2623)

14:46 ISDNSS_A 16.3 5ð 36 734 83 32

15:ð1 ISDNSS_A 16.3 6 24 32 14 52
15:16 ISDNSS_A 16.3 ð ð ð ð ð

Figure 7-73. Resource Interval Report: Communications Line Detail - NWI Maintenance Channel

7-86 Performance Tools V4R2

Communications Line Detail–IDLC Samples
Resource Interval Report ð5/22/96 1ð:29:4ð
Communications Line Detail Page 15
Member . . . : ECL Model/Serial . : 5ðð-2142/1ð-1ðDFD Main storage . . : 32ð.ð M Started . . . : ð4/15/96 1ð:35:3ð
Library . . : PM37CT System name . . : ABSYSTEM Version/Release . : 3/7.ð Stopped . . . : ð4/15/96 12:35:32
PROTOCOL = IDLC (SORT BY INTERVAL)
IOP Transmit/ ---Frames---- ---Frames----
Name/ Receive/ Bytes -Transmitted- Bytes --Received--- Receive Short
Itv Network Line Line Average Trnsmitd Pct Recd Pct CRC Aborts Sequence Frame
End Interface Descriptn Speed Line Util Per Sec Total Err Per Sec Total Err Errors Recd Error Errors
----- ---------- ---------- ----- --------- -------- --------- --- ------- --------- --- ------- ------- -------- -------
CCð5
(26ð5)
11:43 ISDNA IDLCAð1 64.ð ðð/ðð/ðð 42 49 4 33 47 2 ð ð ð ð
11:43 ISDNB IDLCBð1 64.ð ðð/ðð/ðð 2 1 ð ð ð ð ð ð ð ð

Figure 7-74. Resource Interval Report: Communications Line Detail–IDLC

Resource Interval Report ð5/22/96 1ð:29:4ð

Communications Line Detail Page 17
Member . . . : ECL Model/Serial . : 5ðð-2142/1ð-1ðDFD Main storage . . : 32ð.ð M Started . . . : ð4/15/96 1ð:35:3ð
Library . . : PM37CT System name . . : ABSYSTEM Version/Release . : 3/7.ð Stopped . . . : ð4/15/96 12:35:32
PROTOCOL = IDLC (SORT BY INTERVAL)
IOP
Name/
Itv Network Line
End Interface Description Channel
----- ---------- ----------- -------
CCð5
(26ð5)
11:43 ISDNA IDLCAð1 B1
11:43 ISDNB IDLCBð1 B1

Figure 7-75. Resource Interval Report: Communications Line Detail - IDLC

IOP Utilizations–Sample

Chapter 7. Performance Reports—Manager Feature 7-87

| Resource Interval Report ð9/15/98 1ð:54:ð2
| IOP Utilizations Page 45
| Sample Resource Interval Report
| Member . . . : ONE Model/Serial . : 436-21ð6/1ð-1ðBDA Main storage . . : 256.ð M Started . . . : ð9/1ð/98 ðð:16:22
| Library . . : DFLBUGDL1 System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/1ð/98 23:5ð:ð8
| IOP Name/ Itv - IOP Processor Utils - DASD Ops per sec - KB per I/O - KBytes Transmitted Avail Local
| (Model) End Total Comm LWSC DASD Reads Writes Read Write IOP System Storage (K) Util 2
| ----------------- ----- ----- ----- ----- ----- ----------- ----------- ------- ------- ----------- ----------- ----------- ------
| CCð1 (2623) ðð:31 .3 .ð 1 ð 1,73ð .ð
| ðð:46 .3 .ð 1 ð 1,73ð .ð
| ð1:ð1 .3 .ð 1 ð 1,73ð .ð
| ð1:16 .3 .ð 1 ð 1,73ð .ð
| ð1:31 .2 .ð 1 ð 1,73ð .ð
| ð1:46 .3 .ð 1 ð 1,73ð .ð
| ð2:ð1 .2 .ð 1 ð 1,73ð .ð
| ð2:16 .3 .ð 1 ð 1,73ð .ð
| ð2:31 .2 .ð 1 ð 1,73ð .ð
| ð2:46 .2 .ð 1 ð 1,73ð .ð
| ð3:ð1 .2 .ð 1 ð 1,73ð .ð
| ð3:16 .2 .ð 1 ð 1,73ð .ð
| ð3:31 .2 .ð 1 ð 1,73ð .ð
| ð3:46 .2 .ð 1 ð 1,73ð .ð
| ð4:ð2 .2 .ð 1 ð 1,73ð .ð
| ð4:17 .2 .ð 1 ð 1,73ð .ð
| ð4:32 .2 .ð 1 ð 1,73ð .ð
| ð4:47 .3 .ð 1 ð 1,73ð .ð
| ð5:ð2 .2 .ð 1 ð 1,73ð .ð
| ð5:17 .2 .ð 1 ð 1,73ð .ð
| ð5:32 .2 .ð 1 ð 1,73ð .ð
| ð5:47 .3 .ð 1 ð 1,73ð .ð
| ð6:ð2 .2 .ð 1 ð 1,73ð .ð
| ð6:17 .2 .ð 1 ð 1,73ð .ð
| ð6:32 .2 .ð 1 ð 1,73ð .ð
| ð6:47 .2 .ð 1 ð 1,73ð .ð
| IOP Name/ -- Input/Output processor resource name and
| (Model) model number of the attached device
| Itv End -- Interval end time (hour and minute)
| IOP Processor Util Total -- Total utilization for IOP
| IOP Processor Util Comm -- Utilization of IOP due to communications activity
| IOP Processor Util LWSC -- Utilization of IOP due to local workstation activity
| IOP Processor Util DASD -- Utilization of IOP due to DASD activity
| DASD Ops per sec Reads -- Number of reads per second
| DASD Ops per sec Writes -- Number of writes per second
| K Per Read -- Average number of kilobytes (1ð24) per read operation
| K Per Write -- Average number of kilobytes (1ð24) per write operation
| IOP KBytes Transmitted -- Number of Kbytes transmitted from the IOP to the system across the bus
| System KBytes Transmitted-- Number of Kbytes transmitted from the system to the IOP cross the bus
| Avail Local Storage (K) -- Number of kilobytes (1ð24) of local storage that is free
| Util 2 -- Utilization of co-processor

Figure 7-76. Resource Interval Report: IOP Utilizations

Local Work Station Response Times–Sample

7-88 Performance Tools V4R2

 Resource Interval Report ð9/24/98 ð7:4ð:58
Local Work Station Response Times Page 8
Sample Resource Interval Report
Member . . . : TEST2ð Model/Serial . : 5ðð-2142/1ð-317CD Main storage . . : 128.ð M Started . . . : ð9/19/98 16:47:34
Library . . : RWSDATA System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/19/98 17:12:36
IOP Name/ Work Station Itv Active Rsp
(Model) Controller End Util Wrk Stn ð.ð- 1.ð 1.ð- 2.ð 2.ð- 4.ð 4.ð- 8.ð > 8.ð Time
----------------- ------------ ----- ---- ------- --------- --------- --------- --------- ------- ------
WSð1 (916A) 16:52 2.6 ð ð ð ð ð ð .ðð
16:57 2.6 ð ð ð ð ð ð .ðð
17:ð2 2.6 ð ð ð ð ð ð .ðð
17:ð7 2.6 ð ð ð ð ð ð .ðð
17:12 2.7 ð ð ð ð ð ð .ðð
--------- --------- --------- --------- ------- ------
Total Responses: ð ð ð ð ð .ðð
IOP Name/ -- Input/Output processor resource name and
(Model) model number of the attached device
Work Station Controller -- Work station controller description name
Itv End -- Interval end time (hour and minute)
Util -- Percentage of utilization for each IOP
Active Wrk Stn -- Number of work stations with activity
ð.ð- 1.ð -- Number of response times between ð.ð and 1.ð seconds
1.ð- 2.ð -- Number of response times between 1.ð and 2.ð seconds
2.ð- 4.ð -- Number of response times between 2.ð and 4.ð seconds
4.ð- 8.ð -- Number of response times between 4.ð and 8.ð seconds
> 8.ð -- Number of response times > 8.ð seconds
Rsp Time -- Average external response time (in seconds) for
work stations on this controller

Figure 7-77. Resource Interval Report: Local Work Station Response Times

Remote Work Station Response Times–Sample

| Resource Interval Report ð9/24/98 ð7:4ð:58
| Remote Work Station Response Times Page 9
| Sample Resource Interval Report
| Member . . . : TEST2ð Model/Serial . : 5ðð-2142/1ð-317CD Main storage . . : 128.ð M Started . . . : ð9/19/98 16:47:34
| Library . . : RWSDATA System name . . : ABSYSTEM Version/Release . : 4/2.ð Stopped . . . : ð9/19/98 17:12:36
| IOP Name/ Work Station Itv Active Rsp
| (Model) Controller End Wrk Stn ð.ð- 1.ð 1.ð- 2.ð 2.ð- 4.ð 4.ð- 8.ð > 8.ð Time
| ----------------- ------------ ----- ------- --------- --------- --------- --------- ------- ------
| CCð2 ( ) ABSYSTEM 16:52 1 162 ð ð ð ð .ð2
| 16:57 1 174 ð ð ð ð .ð2
| 17:ð2 1 195 ð ð ð ð .ð3
| 17:ð7 2 314 ð ð ð ð .ð2
| ---------- ---------- ---------- ---------- ---------- ------
| Total Responses: 845 ð ð ð ð .ð2
| IOP Name/ -- Input/Output processor resource name and
| (Model) model number of the attached device
| Work Station Controller -- Work station controller description name
| Itv End -- Interval end time (hour and minute)
| Active Wrk Stn -- Number of work stations with activity
| ð.ð- 1.ð -- Number of response times between ð.ð and 1.ð seconds
| 1.ð- 2.ð -- Number of response times between 1.ð and 2.ð seconds
| 2.ð- 4.ð -- Number of response times between 2.ð and 4.ð seconds
| 4.ð- 8.ð -- Number of response times between 4.ð and 8.ð seconds
| > 8.ð -- Number of response times > 8.ð seconds
| Rsp Time -- Average external response time (in seconds) for
| work stations on this controller

Figure 7-78. Resource Interval Report: Remote Work Station Response Times

Batch Job Trace Report

Chapter 7. Performance Reports—Manager Feature 7-89

 Printing the Batch Job Trace Report
Use the Print Trace Report (PRTTRCRPT) command. Before you print the Batch
Job Trace Report, you must use the Start Performance Monitor (STRPFRMON)
command with the JOBTRCITV and JOBTYPE options and the Print Transaction
Report (PRTTNSRPT) command with the *FILE option. The PRTTNSRPT
command creates the QTRJOBT file that the Batch Job Trace Report uses.

What Is the Batch Job Trace Report?

The Batch Job Trace Report shows the progression of different job types (for
example, batch jobs) traced through time. Resources utilized, exceptions, and state
transitions are reported.

See “Performance Report Columns” on page 7-102 for definitions of specific

columns in the reports.

Job Summary
The Job Summary section of the Batch Job Trace Report gives the number of
traces, the number of I/O operations, the number of seize and lock conflicts, the
number of state transitions for each batch job.

See the sample report shown in Figure 7-79.

Job Summary Report–Sample

| Batch Job Trace Report 9/ð5/98 14:15:1ð
| Job Summary Page 1
| Sample Job Trace Report
| Member . . . : Q981421246 Model/Serial . : 5ðð-2142/1ð-18ð3D Main storage . . : 128.ð M Started . . . . : ð5/22/98 12:47:35
| Library . . : THREAD1 System name . . : ABSYSTEM Version/Release : 4/ 2.ð Stopped . . . . : ð5/22/98 12:52:38
| ---- Physical ---- Seize --- State ---
| Job User Job -- Job -- Number CPU ---- I/O Count ---- and Lock --- Transitions ---
| Name Name Number Pool Type Pty Traces Util Sync Async Conflicts A-A A-I
| ---------- ---------- ------ ---- ---- --- ------ ---- ---------- ---------- --------- --------- ---------
| QPFRMON QPGMR ð13842 ð2 B ð 5 11.7 6ð4 235 ð 1 ð
| Job Name -- Name of the job
| User Name -- User name
| Job Number -- Job number
| Pool -- Pool in which the job ran
| Job Type -- Job type and subtype
| Job Pty -- Priority of the job
| Number Traces -- Number of traces
| CPU Util -- Percentage of available CPU time used. This is the average of all processors
| Physical I/O Count
| Sync -- Number of synchronous I/O operations
| Async -- Number of asynchronous I/O operations
| Seize and Lock Conflicts -- Number of seize conflicts and lock waits
| State Transitions A-A -- Number of active-to-active transitions
| State Transitions A-I -- Number of active-to-ineligible transitions

Figure 7-79. Sample of Job Summary Report

Performance Trace Database Files

The Print Transaction Report (PRTTNSRPT) command has options to build for-
matted database files. These files can extend your performance analysis capabili-
ties beyond what the standard trace reports provide.

Using parameters on this command, you can specify a combination of reports and
files to be built in a single run, select specific time ranges and jobs, and limit the

7-90 Performance Tools V4R2

 amount of report and file data produced. If you specify *FILE on the RPTTYPE
parameter of the PRTTNSRPT command, the transaction report creates the files
QTRTSUM, QTRJSUM, and QTRJOBT. If you specify *TRCDTA on the RPTTYPE
parameter of the PRTTNSRPT command, the report creates the file QTRDMPT.

QTRTSUM and QTRJOBT Files

The transaction summary file (QTRTSUM) and job TSE (time slice end)
(QTRJOBT) files have the same format; however, they represent different types of
information.
Ÿ QTRTSUM (transaction summary) file contains one record for every interactive
transaction identified by the PRTTNSRPT command.
Ÿ QTRJOBT (job time slice end) file contains one record per time slice end for all
jobs. Time slice end records are created if the job CPU usage reaches one of
the following values:
– External CPU time slice value
– An internal time slice value defined by the Start Performance Monitor
(STRPFRMON) command.

In the QTRTSUM file, the summary data represents the activity for the transaction.
In the QTRJOBT file, the summary data represents activity that has occurred since
the last TSE or other multiprogramming level trace record.

Table 7-7 (Page 1 of 4). QTRTSUM File

Field Name Description
TRNYEAR Transaction start year
TRNMONTH Transaction start month
TRNDAY Transaction start day
TRNHOUR Transaction start time hour
TRNMIN Transaction start time minute
TRNSEC Transaction start time second
TRNSECD Transaction start time decimal (milliseconds)
TRQUAL Trace qualifier (QTRJOBT file only)
Ÿ 139–Job external time slice end.
Ÿ 145–Job internal time slice end (the CPU time used). The value is specified in the
JOBITVTRC parameter on the STRPFRMON command.
TSKJOB Job name
TSKUSR User name
TSKNUM Task number
TDENUM TDE number (system assigned)
TSPOOL Main storage pool in which the job ran
TPRTY Current job priority
TTYPE Job type and subtype.
Refer to explanation of the Typ field in “Job Summary Section” on page 7-38 for a list of
types and subtypes.

Chapter 7. Performance Reports—Manager Feature 7-91

 Table 7-7 (Page 2 of 4). QTRTSUM File
Field Name Description
TPURGE Purge attribute (Y/N). Defines whether or not the job is eligible to have its PAG purged at
the end of the transaction.
TRSP Response time (in seconds). The time from the first W→A transition to the last A→W tran-
sition in the transaction.
TCPU CPU time (in seconds) used by this job during the transaction. It does not include the CPU
time for asynchronous server tasks such as Licensed Internal Code work station IOM,
asynchronous disk I/O tasks, and others
TSDBRD Synchronous database reads (count)
TSDBWRT Synchronous database writes (count)
TSNDBRD Synchronous nondatabase reads (count)
TSNDBWRT Synchronous nondatabase writes (count)
TADBRD Asynchronous database reads (count)
Refer to the IOPND and SYSYNC fields in the QAITMON file created by the
WRKSYSACT command or to the JBIPF and JBIOW fields in the QAPMJOBS sample
data file created by the STRPFRMON command for the job to see how many asynchro-
nous disk reads were turned into synchronous reads.
TADBWRT Asynchronous database writes (count)
TANDBRD Asynchronous nondatabase reads (count) (See field TADBRD.)
TANDBWRT Asynchronous nondatabase writes (count)
TPAGFLT Process access group (PAG) faults (count)
TBIN Binary overflow count
TDEC Decimal overflow count
TEAOCNT Reserved
TCHKSUM Reserved
TACT Time in the activity level (in seconds)
TWAIT Short wait time in the activity level (in seconds)
TINELW Wait-to-ineligible (W→I) transition time waiting for activity level (in seconds). This occurs
after coming out of a long wait such as start of transaction or the end of a lock wait.
TINELA Active-to-ineligible (A→I) time slice end (TSE) transition time waiting for activity level (in
seconds). This occurs after leaving the activity level at external time slice end because
other equal or higher priority jobs were waiting for an activity level.
TAICNT The number of active-to-ineligible (A→I) transitions. The number of external time slice
ends that caused the job to leave the activity level because there were equal or higher
priority jobs waiting for an activity level.
TAACNT The number of active-to-active (A→A) transitions. The number of external time slice ends
that did not cause the job to leave the activity level because there were no equal or higher
priority jobs waiting for an activity level.

7-92 Performance Tools V4R2

Table 7-7 (Page 3 of 4). QTRTSUM File
Field Name Description
TEXWTM Total exceptional wait time (in seconds). Sum of the following fields:
TINELA Active-ineligible wait
TWAIT Short wait
TSWXTM Short wait extended
TSZTM Seize wait
TLCKTM Lock wait
T3270 3270 wait
TDDM DDM wait
TEVTM Event wait
TXATM Total excess active time (added only for interactive jobs)
TDELTM Delay time (added only for noninteractive jobs)
no-name Miscellaneous wait time. For example save/restore, diskette, or tape mount and
respond to mount message.
TSWTM Total short wait time (in seconds). Short wait time (SW time in the transaction reports) is
the time spent waiting for an event (such as work station output complete) while remaining
in the active state. When the short wait ends (it ends automatically after 2 seconds), the
job goes into the short wait extended state.
TSWXTM Total short wait extended (short wait time-out) time (in seconds). See field TSWTM. During
the time a job is in short wait extended (abbreviated SWX in the transaction reports), it
does not hold an activity level (it is in the wait state). The short wait is satisfied when the
waited-on event occurs.
TSWXCNT Total number of short waits extended. The number of short waits where the job was taken
out of the activity level after 2 seconds and put into long wait (an A→W transition).
TSZTM Total seize conflict wait time (in seconds). Total time this job waited in the activity level for
seize conflicts.
TLCKTM Total lock conflict wait time (in seconds). Total time this job waited outside the activity
level for lock conflicts.
THOLDTM Total seize/lock conflict hold time to other jobs. The field contains the total time that other
jobs waited for objects held by this job. For example, when a job held an object for 2
seconds and for that time two other jobs waited for the object, the THOLDTM value would
be 4.
TEVTM Total event wait time (in seconds)
TXATM Total excess active time (in seconds). A calculated value, not a measured value, that
represents the time a job was in the activity level, could not use the CPU, and is not
accounted for by other measurements. It can be the result of waiting behind equal or
higher priority jobs for the CPU, waiting to do disk I/O, or waiting for an internal, non-
instrumented object such as the free space lock on storage management.
T3270 Total 3270 emulation wait time
TDDM Total DDM wait time (in seconds)
TMRT Total MRT wait time (in seconds)
TDELTM Total long wait time (in seconds) such as key/think wait time or delay time
TSZLCKCT Seize and lock conflicts encountered by this job (count)
TSZLCKRL Seize and lock releases done by this job when other jobs waited (count)
TBMPL Number of jobs holding an activity level in this job's pool at transaction start (count)

Chapter 7. Performance Reports—Manager Feature 7-93

 Table 7-7 (Page 4 of 4). QTRTSUM File
Field Name Description
TIMPL Number of jobs waiting for activity level in this job's pool at transaction start (count)
TPGM1 First program name in stack at the end of transaction
TPGM2 Second program name in stack at the end of transaction
TPGM3 Third program name in stack at the end of transaction
TPGM4 Fourth program name in stack at the end of transaction
TPGM Program that caused transaction (one of these field names is the application in control of
the transaction)
TELAP Elapsed time of transaction (in seconds)
TPVPGM Previous program name
TIPRTY Assigned job priority
| TTHID Thread identifier
| TTHFLG Secondary thread flag (0=primary thread, 1=secondary thread)

QTRJSUM File
The job summary file QTRJSUM contains one record for each job or task listed on
the PRTTNSRPT job summary report.

Table 7-8. QTRJSUM File

Field Name Description
TDENUM TDE number. Licensed Internal Code task dispatching element.
JOBID Job name or user name
| USERID User ID
JOBNUM Job number
POOL Storage pool the job started in
JOBTYP Job type code
Refer to explanation of the Typ field in “Job Summary Section” on page 7-38 for a list of
types and subtypes.
TRCPER Trace period sequence number (reserved)
JDATE Job start date MM/DD/YY
JSTART Job start time HH:MM:SS
JSTOP Job stop time HH:MM:SS
JELAP Job total elapsed time (in seconds)
JCPU Job total CPU time (in seconds)
JDBIO Job total disk database reads
JNDBIO Job total disk nondatabase reads
JWRTIO Job total disk writes

7-94 Performance Tools V4R2

 The following fields are totals for the duration of the job, unless the beginning or end time selection option was
taken on the PRTTNSRPT command.
Then the values are for the selected time only.

TRNNUM Total number of transactions (type I jobs only)

JBEG Transaction report selection beginning time HHMMSS
JEND Transaction report selection ending time HHMMSS
JOBELP Total elapsed time (in seconds) for the job (job start to job end)
JOBCPU CPU time (in seconds) the job used
JOBDB Total disk database reads
JOBNDB Total disk nondatabase reads
JOBWRT Total disk writes
JARSP Average transaction response time (in seconds) (type I jobs only)
JMRSP Maximum transaction response time (in seconds) (type I jobs only)
JACPU Average CPU time per transaction (in seconds) (type I jobs only)
JMCPU Maximum CPU time by a transaction (in seconds) (type I jobs only)
JADBR Average disk database reads per transaction (type I jobs only)
JANDBR Average disk nondatabase reads per transaction (type I jobs only)
JAWRT Average disk writes per transaction (type I jobs only)
JAIO Average disk I/O per transaction (type I jobs only)
JMIO Maximum disk I/O by a transaction (type I jobs only)
JWI Total number of W→I transitions
JAI Total number of A→I transitions
JLCKS Total number of lock conflicts
JATM Total time the job was in an activity level (in seconds)
JWTM Total short wait time in an activity level (in seconds)
JINELW Total ineligible time as a result of wait-to-ineligible transitions (in seconds)
JINELA Total ineligible time as a result of active-to-ineligible transitions (in seconds)
JLKWTM Total wait time for short wait and short wait extended, QEM wait, DDM wait, and
save/restore, diskette, or tape wait.
JKYTK Total key/think time (in seconds)
TSKID Combined job name, user ID, and user name fields
JSPRTY Assigned job priority
| JTHID Thread identifier
| JTHFLG Secondary thread flag (0=primary thread, 1=secondary thread)

Chapter 7. Performance Reports—Manager Feature 7-95

QTRDMPT File
The QTRDMPT file is a version of the QAPMDMPT file formatted as a database
file. It gives you access to each performance monitor trace record created. The
QAPMDMPT file is built when the performance monitor ends, tracing is specified,
and the monitor is directed to dump the internal trace data to a database file. The
QAPMDMPT file can also be built with the DMPTRC command after the monitor
shuts down.

The field names shown below without asterisks contain information taken directly
from the QAPMDMPT file. Field names shown below with an asterisk (*) in front of
them contain information created by the transaction report. Unless otherwise speci-
fied, numeric values are in decimal.

Table 7-9 (Page 1 of 4). QTRDMPT File

Field Name Description
DSEQNM Sequence number in QAPMDMPT (relative record in file)
DTID Trace ID in hexadecimal
X'68' Resource management (seize/lock activity)
X'70' MPL trace record (job state transitions)
X'73' Trace control record (job/task start/stop/existence)
X'AB' Transaction boundary trace record
X'AC' Transaction boundary trace record–source pass-through, target pass-through,
and WSF (work station function) target pass-through

All other trace identifiers are ignored by the transaction report.

7-96 Performance Tools V4R2

Table 7-9 (Page 2 of 4). QTRDMPT File
Field Name Description
DTQUAL Trace qualifier
If DTID = X'68' seize/lock trace (Licensed Internal Code tasks or OS/400 jobs), the trace
qualifiers are:
701 Job/task entered seize conflict wait
1001 Job/task released seize that is being waited on
903 Job entered lock conflict wait
906 Job released lock that is being waited on
If DTID = X'70' MPL trace (OS/400 jobs only), the valid trace qualifiers for active state
codes are:
129 Ineligible-to-active transition
131 Message received and job was in current activity level when the message was
received
133 Dequeue after time-out; job in current activity level when the message was
received
135 Wait-to-active
137 Wait timed out, no message received; wait-to-active
139 Active-to-active (job external time slice end)
142 Wait-to-active (job is already in activity level)
145 STRPFRMON pseudo TSE; active-to-active
The trace qualifiers for wait state codes are:
128 Just initiated job cannot get into activity level
130 Active-to-wait transition; drop from activity level
132 Wait-to-ineligible transition
134 Active-to-wait but stay in activity level
136 Time slice end; active-to-ineligible
If DTID = X'73' control trace (OS/400 jobs and SLIC tasks), the valid qualifiers are:
130 Job started while trace was active
133 Job ended while trace was active
127 Job active at start of trace
136 Job active at end of trace
129 SLIC task started while trace was active
132 SLIC task ended while trace was active
126 SLIC task active at start of trace
135 SLIC task active at end of trace
DTRDAT Transition date YYYYMMDD
DTRTM Transition time HHMMSSmmm
DTRHR Transition hour xx.xxxxxxx
DTRELP Elapsed seconds from previous state
*DPVDAT Previous transition date YYYYMMDD
*DPVTM Previous transition time HHMMSSmmm
*DPVHR Previous transition hour xx.xxxxxxx
DTDEHX TDE number in hexadecimal
DSPOOL Pool number in which job ran
DPRTY Current job priority
DTYPE Job type and subtype.
Refer to explanation of the Typ field in “Job Summary Section” on page 7-38 for a list of
types and subtypes.

Chapter 7. Performance Reports—Manager Feature 7-97

 Table 7-9 (Page 3 of 4). QTRDMPT File
Field Name Description
DPURGE Job purge attribute: 0=No, 1=Yes
DCPU Total CPU time (in seconds)
*DLPCPU CPU time since last job state transition (in seconds)
*DCPUPC Percentage of CPU usage since last job state transition
DJOBNM Job name
DUSRNM User name
DJOBNB Job number
*DTRSTA Transition to this state; this matches what is shown in the Transition Report
*DTRWAT Transition wait code; this matches what is shown in the Transition Report
DMPL Current number of pool activity levels in use
DIPL Number of ineligible jobs waiting for pool activity level
DCSDR Synchronous database reads (cumulative)
*DISDR Synchronous database reads (since last transition)
DCSDW Synchronous database writes (cumulative)
*DISDW Synchronous database writes (since last transition)
DCSNR Synchronous nondatabase reads (cumulative)
*DISNR Synchronous nondatabase reads (since last transition)
DCSNW Synchronous nondatabase writes (cumulative)
*DISNW Synchronous nondatabase writes (since last transition)
DCADR Asynchronous database reads (cumulative)
*DIADR Asynchronous database reads (since last transition)
DCADW Asynchronous database writes (cumulative)
*DIADW Asynchronous database writes (since last transition)
DCANR Asynchronous nondatabase reads (cumulative)
*DIANR Asynchronous nondatabase reads (since last transition)
DCANW Asynchronous nondatabase writes (cumulative)
*DIANW Asynchronous nondatabase writes (since last transition)
DCPAG Process access group (PAG) faults (cumulative)
*DIPAG Process access group (PAG) faults (since last transition)
DCEAO Reserved
*DIEAO Reserved
DCCKSM Reserved
*DICKSM Reserved
DCDEC Decimal overflow exceptions (cumulative)
*DIDEC Decimal overflow exceptions (since last transition)
DCBIN Binary overflow exceptions (cumulative)
*DIBIN Binary overflow exceptions (since last transition)
DCFLP Floating point overflow exceptions (cumulative)

7-98 Performance Tools V4R2

Table 7-9 (Page 4 of 4). QTRDMPT File
Field Name Description
*DIFLP Floating point overflow exceptions (since last transition)
DCPWT Permanent writes (cumulative)
*DIPWT Permanent writes (since last transition)
DPGM1 Program 1 (last) (DTID = X'70' only)
DPGM2 Program 2 (second from the last) (DTID = X'70' only)
DPGM3 Program 3 (third from the last) (DTID = X'70' only)
DPGM4 Program 4 (fourth from the last) (DTID = X'70' only)

Resource management data. The following three fields contain valid information only for records that have
DTID=X'68' (Resource Management Trace).

DSLJOB Job/task name of seize/lock waiter/holder

DSLUSR User name of seize/lock waiter/holder
DSLNBR Job number of seize/lock waiter/holder

The following five fields can have data that is not valid if the object was destroyed before the trace was dumped
to the QAPMDMPT file.

DSLOTY Seize/lock object type

Note: Object types and codes can be found in AS/400 Licensed Internal Code Diagnostic
Aids - Volume 1.
DSLOLB Seize/lock object library name (may be undefined for machine objects) A machine object
is a program object that has no defined storage form; the object is defined internally to the
machine.
DSLOFL Seize/lock object file/object name (may be undefined for machine objects)
DSLOMB Seize/lock object member name (database files only)
DSLRRN Relative record number of the lock database file (if report is run on same system that it
was collected on and the file still exists)
DRSVD1 Reserved
DRSVD2 Reserved

Transaction boundary information. These fields contain valid information only for trace records with DTID = X'AB'
or X'AC'.

Chapter 7. Performance Reports—Manager Feature 7-99

 DTNTY Transaction type (in decimal)
1 Display I/O
2 Data queue
3 MRT
4 Source pass-through
5 Target pass-through
6 WSF target pass-through
DTNSTY Transaction subtype (in decimal)
If DTNTY = 1, 2, or 3:
1 Start transaction
2 End transaction
3 End response time transaction (for DTNTY = 1 only)
IF DTNTY = 4, 5, or 6:
1 Start transaction
2 End transaction
3 Start session
4 End session
DTNBIT Reserved
DTNNM1 Name of display device for display I/O transactions
Name of data queue library for data queue transactions
Name of display device for MRT transactions
Name of device description for pass-through transactions
DTNNM2 Name of display file for display I/O transactions
Name of data queue for data queue transactions
Name of display file for MRT transactions
Name of target control point for source pass-through
transactions
Name of source control point for target pass-through
transactions
Name of controller description for WSF target
pass-through transactions
DTNNM3 Reserved
DTNNM4 Reserved
DTNDAT Date of transaction YYYYMMDD
DTNTM Time of transaction HHMMSSmmm
DTNHR Hour of transaction xx.xxxxxxx
*DTNBDY Transaction boundary flag:
Set to 1 if this trace record is at a transaction boundary; set to 0 if it is not at a transaction
boundary.
DTNID Reserved
DIPRTY Assigned job priority
| DTHID Thread identifier
| DTHFLG Secondary thread flag (0=primary thread, 1=secondary thread)

7-100 Performance Tools V4R2

 QAPTLCKD File
QAPTLCKD is the file created by using the Print Lock Report (PRTLCKRPT) from
information in the QAPMDMPT trace data file. This file contains data on seizes and
locks. Table 7-10 shows a description of each field in the QAPTLCKD file.

Table 7-10 (Page 1 of 2). QAPTLCKD File

Field Name Description
SLWTOD Time of day (HH.MM.SS) that the requesting job REQNAM had either a seize or lock con-
flict on the object OBJNAM that was held by job HLDNAM.
SLWLEN Length of time (in milliseconds) from the start of the object conflict until the holding job
released the object. This is not necessarily the amount of time that the requesting job is
delayed in getting the object. That time may be longer than the conflict delay.
SLCODE The type of conflict: blank = Seize, L = Lock. Seizes occur only in Licensed Internal Code
or implicitly within high-level MI instructions such as Create Object or Add to a File. Locks
occur in jobs running in the OS/400 program and can be explicitly requested.
REQTDE Requesting job's TDE number
REQNAM Requesting job's name, user ID, job number
Position Value
1-10 Job name
12-21 User name
23-28 Job number
| REQTTH Requesting job's thread identifier
HLDTDE Holding job's TDE number
HLDNAM Holding job's name, user ID, job number
Position Value
1-10 Job name
12-21 User name
23-28 Job number
| HLDTTH Holding job's thread identifier
OBJADR The address of the object
Position Value
1-8 Segment address
9-12 Offset
Note: AS/400 Licensed Internal Code Diagnostic Aids - Volume 1 contains a description
of the system's preassigned addresses.

Chapter 7. Performance Reports—Manager Feature 7-101

 Table 7-10 (Page 2 of 2). QAPTLCKD File
Field Name Description
OBJNAM The object type, name, library (if applicable), and member (if applicable)
Position Value
1-6 Object type description
8-17 Name
19-28 Library
30-39 Member (database files and indexes)
In some cases the object type may not be translated; instead it may be a 2-byte
hexadecimal code. If the object name is not meaningful, it is possible that the object
address is one of the system's preassigned addresses. AS/400 Licensed Internal Code
Diagnostic Aids - Volume 1contains more information on these objects and codes.
OBJRRN Database file record number. Valid only for type DS (data space) when the Print Lock
Report (PRTLCKRPT) command created the QAPTLCKD file on the same system that the
data was collected on.

Performance Report Columns

>8.0 (Component) The number of times the response time was greater than 8
seconds.
---------- (pgmname) (Transaction) The transaction totals record. For example,
---------- QUYLIST, as shown in Figure 7-49 on page 7-62. This report
line occurs each time the job has an active-to-wait transaction. Totals
are created for Rsp* (response time), CPU Secs, and I/O counts for the
transaction.
A-I Wait /Tns (Transaction) The average time, in seconds, of active-to-ineligible
wait time per transaction. If this value is high, it may be because the
time-slice value is set too low for many of the interactive jobs. Consider
increasing the time slice-value.
Aborts Recd (Resource Interval) The number of frames received that contained
HDLC abort indicators. This indicates that the remote equipment ended
frames before they were complete.
Act Jobs (Job Interval) The number of selected jobs (interactive or noninteractive,
depending on the report section) that were active during the interval.
Act Level (Component) Initial pool activity level.
Act Lvl (System, Pool Interval) Activity level. For the Pool Activity section of the
Pool Interval Report, the activity level of the pool during the interval. For
the Storage Pool Utilization section of the System Report, the activity
level at the time of the first sample interval.
Act-Inel (System, Component) Average number of active-to-ineligible job state
transitions per minute.
Act-Wait (System, Component) Number of transitions per minute from active state
to wait state by processes assigned to this pool.
Active Devices (System) Average number of active devices on the line.

7-102 Performance Tools V4R2

Active display stations (local or remote) (System) The number of local or remote
display stations entering transactions during the measurement period.
Active Jobs (Transaction) The number of interactive jobs that were active during
the interval.
Active Jobs Per Interval (System) Average number of jobs of this type that were
active per sample interval.
Active K/T /Tns (Transaction) An average think time and keying time (or the delay
time between the end of one transaction and the start of the next trans-
action), in seconds, for the active work stations (described under Est of
AWS). Active K/T /TNS delay time differs from Key/Think /TNS delay
time in that any delay time greater than 600 seconds has been rounded
to 600 seconds. This technique is used to reduce the effect of very
casual users (those who may do intermittent work or leave their work
stations for long periods of time) on the estimate of active work stations.
Active Wrk Stn (Resource Interval) The number of work stations with activity.
Active/Rsp (Transaction) The time the job spends (either waiting or active) during
transaction processing, while it holds an activity level.
Activity level (System) The sum of activity levels for all interactive pools that had
interactive job activity running in them.
Activity Level Time (Transaction) A breakdown of the transaction time spent
ACTIVE, waiting on a SHORT WAIT, and waiting on a SEIZE/CFT
(seize conflict). The SHORT WAIT and SEIZE CFT time are included
under ACTIVITY LEVEL TIME, because the activity-level slot is not
given up during these times. Note that the seize conflict time is included
in the active time, not added to it to get transaction/response time, as is
the case for waiting time.
Arith Ovrflw (Component, Job Interval) The number of arithmetic overflow
exceptions that occurred for the selected interactive jobs during the
interval.
ASP ID (System, Resource Interval) Auxiliary storage pool identifier.
Async (System, Component, Transaction, Job Interval) The number of asyn-
chronous disk I/O operations started by the selected interactive jobs
during the interval. The job that starts the I/O operation may continue
processing without having to wait for the I/O operation to complete. The
I/O operation is completed by a background system test.
Async DIO /Tns (Transaction) The sum of the averages of the asynchronous DB
READ, DB WRITE, NDB READ, and NDB WRITE requests (the average
number of asynchronous I/O requests per transaction for the job).
Async Disk I/O (System, Component, Transaction) Number of asynchronous disk
input/output operations per transaction.
Async Disk I/O per Second (Component) Average asynchronous disk I/O oper-
ations per second.
Async Disk I/O Requests (Transaction) The total number of asynchronous disk I/O
requests for the given combination of priority, job type, and pool.
Async I/O /Sec (Job Interval) The average number of asynchronous disk I/O oper-
ations started per second by the job during the interval. This is calcu-
lated by dividing the asynchronous disk I/O count by the elapsed time.

Chapter 7. Performance Reports—Manager Feature 7-103

 Async I/O Per Second (Job Interval) The average number of asynchronous disk
I/O operations started per second by the selected noninteractive jobs
during the interval.
Async Max (Transaction) Listed under Average DIO/Transaction, the maximum
number of asynchronous DBR, NDBR, and WRT I/O requests encount-
ered for any single transaction by that job. If the job is not an interactive
or autostart job type, the total disk I/O for the job is listed here.
Async Sum (Transaction) Listed under Average DIO/Transaction, the sum of the
averages of the asynchronous DBR, NDBR, and WRT requests (the
average number of asynchronous I/O requests per transaction for the
job).
Asynchronous DBR (System, Job Interval, Pool Interval) The average number of
asynchronous database read operations on the disk per transaction for
the job during the intervals. This is calculated by dividing the asynchro-
nous database read count by the transactions processed. This field is
not printed if the jobs in the system did not process any transactions.
For the Resource Utilization section of the System Report, it is the
number of asynchronous database read operations per second.
Note: The asynchronous I/O operations are performed by system asyn-
chronous I/O tasks.
Asynchronous DBW (System, Job Interval) The average number of asynchronous
database write operations on the disk per transaction for the selected
jobs during the interval. This is calculated by dividing the asynchronous
database write count by the transactions processed. This field is not
printed if the jobs in the system did not process any transactions. For
the Resource Utilization section of the System Report, it is the number
of asynchronous database read operations per second.
Note: The asynchronous I/O operations are performed by system asyn-
chronous I/O tasks.
Asynchronous disk I/O per transaction (System) The average number of asyn-
chronous physical disk I/O operations per interactive transaction.
Asynchronous NDBR (System, Job Interval, Pool Interval) The average number of
asynchronous nondatabase read operations per transaction for the jobs
in the system during the interval. This is calculated from the asynchro-
nous nondatabase read count divided by the transactions processed.
This field is not printed if the jobs in the system did not process any
transactions. For the Resource Utilization section of the System Report,
it is the asynchronous nondatabase read operations per second.
Note: The asynchronous I/O operations are performed by system asyn-
chronous I/O tasks.
Asynchronous NDBW (System, Job Interval, Pool Interval) The average number
of asynchronous nondatabase write operations per transaction for the
jobs in the system during the interval. This is calculated from the asyn-
chronous nondatabase write count divided by the transactions proc-
essed. This field is not printed if the jobs in the system did not process
any transactions. For the Resource Utilization section of the System
Report, it is the number of asynchronous nondatabase write operations
per second.

7-104 Performance Tools V4R2

 Note: The asynchronous I/O operations are performed by system asyn-
chronous I/O tasks.
Aut Lookup (Component) Number of authority lookups per second. An authority
lookup is the process whereby the Licensed Internal Code determines
whether a particular user ID is authorized to access a specific object.
| Beginning in V3R7, reduced instruction set computer (RISC) systems
| store the most recent private authority lookup results in an authorization
| lookup cache. If the next lookup is for one of the authorities stored in the
| cache, the private authority lookup has a minimal performance degrada-
| tion over public or owner authority performance.
| The cache can store up to 32 private authorities for objects and for
| authorization lists. Whenever the system looks for a private authority,
| the system queries the cache. Whenever an authority is granted or
| revoked for a user, the cache is updated. Performing an IPL clears the
| cache.
| The performance monitor counts each authority lookup. The advisor
| function and the redbook, AS/400 Performance Management
| V3R6/V3R7, SG24-4735, provide CPU utilization estimates based on the
| number of lookups per second and the processor rating. Because of this
| caching capability, the counts are incremented as if they were not
| cached. Therefore, beginning with V3R7, the effect on the CPU utiliza-
| tion counts could be much less than the advisor message and the
| redbook would indicate.
Avail Local Storage (K) (Resource Interval) The number of kilobytes of free local
storage in the IOP.
Available Storage (Component) Available local storage (in bytes). The average
number of bytes of available main storage in the IOP. The free local
storage is probably not joined because it has broken into small pieces.
Average (Transaction) The average value of the item described in the column for
all transactions.
Average Disk Activity Per Hour (Component) See Disk Arm Seek Distance
Average DIO/Transaction (Transaction) Seven columns of information about phys-
ical disk I/O counts. Physical I/O contrasts with logical I/O shown else-
where in these reports. A logical I/O is a request sent at the program
level that might result in an access to auxiliary storage (DASD). A phys-
ical I/O refers to those requests that actually result in access to auxiliary
storage.
Ÿ Synchronous DBR
Ÿ Synchronous NDBR
Ÿ Synchronous Wrt
Ÿ Synchronous Sum
Ÿ Synchronous Max
Ÿ Async Sum
Ÿ Async Max
Average K per I/O (Resource Interval) The average number of kilobytes trans-
ferred during each disk read or write operation.

Chapter 7. Performance Reports—Manager Feature 7-105

 Average Phys I/O /Sec (Resource Interval) The average number of physical disk
read and write operations per second made on all disks on the system.
Average Reads/Sec (Resource Interval) The average number of physical disk read
operations per second made on all disks on the system.
Average Response (System) Average response time (in seconds) for interactive
transactions. The Total/Average interactive response time does not
include transactions for DDM server jobs.
Average Response Time (System) Average disk response time per I/O operation.
Average Response Time (seconds) (System) The average interactive response
time.
Average Service Time (System) Average disk service time per I/O operation. This
is the amount of time a request would take if there were no contention.
Average Wait Time (System) Average disk wait time per I/O operation. Normally
due to contention.
Average Writes/Sec (Resource Interval) The average number of physical disk write
operations per second made on all disks on the system.
Avg CPU /Tns (Transaction) The average number of processing unit seconds per
transaction that fell in the given category.
Avg K/T /Tns (Transaction) The average think time and keying time (or the delay
time between transaction boundaries), in seconds, for the interactive
jobs.
Avg Length (Lock) The average number of milliseconds a lock or seize was held.
Avg Rsp (Sec) (Transaction) The average transaction response time in seconds.
Avg Rsp /Tns (Transaction) The average response per transaction (in seconds) for
the transactions that fell into the given category.
Avg Rsp Time (Component) Average transaction response time.
Avg Sec Locks (Transaction) The average length of a lock in seconds attributed to
interactive or noninteractive waiters.
Avg Sec Seizes (Transaction) The average length of a seize in seconds attributed
to interactive or noninteractive waiters.
Avg Time per Service (Resource Interval) The amount of time a disk arm uses to
process a given request.
Avg Util (System, Resource Interval) On the Disk Utilization Summary of the
Resource Report, the average percentage of available time that disks
were busy. It is a composite average for all disks on the system. On the
Communications Summary of the System Report, the average per-
centage of line capacity used during the measured time interval.
Batch asynchronous I/O per second (System) The average number of asynchro-
nous physical disk I/O operations per second of batch processing.
Batch CPU seconds per I/O (System) The average number of system processing
unit seconds used by all batch jobs for each I/O performed by a batch
job.

7-106 Performance Tools V4R2

Batch CPU Utilization (Component) Percentage of available CPU time used by the
following types of jobs:
Ÿ Batch
Ÿ Autostart
Ÿ Evoke
Ÿ SCPF (Start CPF), spool reader/writer
Note: For a multiple-processor system, this is the average use across
all processors.
Batch impact factor (System) Batch workload adjustment for modeling purposes.
Batch permanent writes per second (System) The average number of permanent
write operations per second of batch processing.
Batch synchronous I/O per second (System) The average number of synchro-
nous physical disk I/O operations per second of batch processing.
BCPU / Synchronous DIO (Transaction) The average number of batch processor
unit seconds per synchronous disk I/O operation.
Bin (Transaction) The number of binary overflow exceptions.
Binary Overflow (Component) Number of binary overflows per second.
BMPL - Cur and Inl (Transaction) The number of jobs currently in the activity level
(beginning current multiprogramming level), and the number of jobs on
the ineligible queue (beginning ineligible multiprogramming level) for the
storage pool that the job ran in when the job left the wait state (the
beginning of the transaction).
Note: Multiprogramming level (MPL) is used interchangeably with
activity level.
Bundle Writes System (Component) Number of bundle writes to internal system
journals. A bundle write is a group of journal entries which are deposited
together by the system.
Bundle Writes User (Component) Number of bundle writes to user-created jour-
nals. A bundle write is a group of journal entries which are deposited
together by the system.
Bytes per Second Received (System) Average number of bytes received per
second.
Bytes per Second Transmitted (System) Average number of bytes transmitted per
second.
Bytes Recd per Sec (Resource Interval) The average number of bytes received
per second.
Bytes Trnsmitd per Sec (Resource Interval) The average number of bytes trans-
mitted per second.
Category (Transaction) A group of transactions categorized together. In the Anal-
ysis by Interactive Transaction Category, the transactions are categor-
ized by the processing unit model. The boundary values that are used to
separate the transactions are given in the Avg CPU /Tns column. For
the Analysis by Interactive Response Time, they are categorized by their
response time. For the Analysis by Interactive Key/Think Time, they are
categorized by their key/think time.

Chapter 7. Performance Reports—Manager Feature 7-107

 Cache Hit Statistics (Component) Statistics data about use of cache including:
Ÿ The percent of Device Cache Read Hit for each arm.
Ÿ The percent of Controller Cache Read Hit for each arm.
Ÿ The percent of efficiency of write cache
Device read Device Read is the number of Device Cache Read Hits
(DSDCRH) divided by number of Device Read Operations
(DSDROP), expressed as a percent
Controller read Controller Read is the number Controller Cache Read
Hits (DSCCRH) divided by number of Read Commands
(DSRDS), expressed as a percent.
Write efficiency Write efficiency is the difference between Write Com-
mands (DSWRTS) and Device Write Operations (DSDWOP)
divided by Write Commands (DSWRTS), expressed as a
percent.
Channel (Resource Interval) The B-channel used by the IDLC line. (special condi-
tion)
Cmn (Job Interval) The number of communications I/O operations performed
by the selected interactive jobs during the interval.
Cmn I/O (Component) Number of communications operations (Get, Put).
Cmn I/O Per Second (Job Interval) The average number of communications I/O
operations performed per second by the selected noninteractive jobs
during the interval.
Collision Detect (Resource Interval) The number of times that the terminal equip-
ment (TE) detected that its transmitted frame had been corrupted by
another TE attempting to use the same bus.
Communications I/O Count (System) Number of communications I/O operations.
Communications I/O Get (System) Number of communication get operations per
transaction.
Communications I/O Put (System) Number of communication put operations per
transaction.
Communications Lines (System, Component, Job Interval, Pool Interval) For the
Report Selection Criteria, the list of communications lines selected to be
included (SLTLINE parameter) or excluded (OMTLINE parameter).
These are the communications line names you specify.
Control Units (System, Component, Job Interval, Pool Interval) The list of control
units selected to be included (SLTCTL parameter) or excluded
(OMTCTL parameter). These are the controller names you specify.
Count (Transaction, Lock) The number of occurrences of the item in the
column. For example, in a lock report, it is the number of locks or
seizes that occurred.
CPU (Transaction) The total processing unit seconds used by the jobs with a
given priority.
CPU /Tns (Transaction, Job Interval) The amount of available processing unit time
per transaction in seconds.

7-108 Performance Tools V4R2

CPU Model (System) The processing unit model number.
CPU per I/O Async (System) CPU use per asynchronous I/O.
CPU per I/O Sync (System) CPU use per synchronous I/O.
CPU per Logical I/O (System) Processing unit time used for each logical disk I/O
operation.
CPU QM (Transaction) The simple processing unit queuing multiplier.
CPU Sec (Transaction) The processing unit time used by the job in this state.
CPU Sec /Sync DIO (Transaction) The ratio of CPU seconds divided by synchro-
nous disk I/O requests for each type of job.
CPU Sec Avg and Max (Transaction) The average processing unit time per trans-
action for the job and the largest processing unit time used for a trans-
action in the job. If the job is not an interactive or autostart job type,
then only the total processing unit time for the job is listed under the
MAX column heading.
CPU Sec per Tns (Transaction) The processing unit time per transaction.
CPU Seconds (System, Transaction, Component) Average processing unit
seconds used per transaction. For System Summary Data, it is the total
available processing unit time used by the jobs during the trace period.
For Priority-Jobtype-Pool Statistics, it is the total processing unit
seconds used by the jobs with a given combination of priority, job type,
and pool. For Batch Job Analysis, it is the amount of available processor
unit time used by the job in seconds. For Concurrent Batch Job Statis-
tics, it is the amount of available processor unit time used by the jobs in
the job set in seconds.
CPU seconds per transaction (System) The average processing unit seconds per
transaction.
CPU Util (System, Component, Transaction, Job Interval, Pool Interval, Batch Job
Trace) Percentage of available processing unit time used. For multiple-
processor systems, this is the total utilization divided by the number of
processors.
CPU Util per Transaction (Component) The result of the CPU Utilization divided
by the total number of transactions for the job.
CPU/Async I/O (Job Interval) The average number of milliseconds of processing
unit time taken for each asynchronous disk I/O operation. This is calcu-
lated by dividing the milliseconds of the processing unit time the job
used by the asynchronous disk I/O count.
CPU/Sync I/O (Job Interval) The average number of milliseconds of processing unit
time taken for each synchronous disk I/O operation. This is calculated
from the milliseconds of the processing unit time used by the job divided
by the synchronous disk I/O count.
CPU/Tns (Transaction) The average number of processing seconds per trans-
action for the job during the interval. This is calculated from the amount
of processing unit time used divided by the number of transactions proc-
essed.
Cpu/Tns (Sec) (Transaction) The number of processing unit seconds per trans-
action.

Chapter 7. Performance Reports—Manager Feature 7-109

 Ctl (Component) Controller identifier.
Cum CPU Util (Transaction) The cumulative percentage of available processing
unit time used by the transactions that have an average response time
per transaction equal to or less than the given category. For example, in
CPU by Priority for All Jobs for Total Trace Period (System Summary
Data), it is the unit time used by the jobs with a priority higher or equal
to the given priority.
Cum Pct Tns (Transaction) Cumulative CPU percent per transaction. For system
summary data, it is the cumulative CPU percentage of all transactions
that have an average response time per transaction equal to or less
than the given category. For Interactive Program Transactions Statistics,
it is the cumulative CPU percentage of all transactions through the listed
program. For Job Statistics section, it is the cumulative CPU percentage
of total transactions through the listed job. For Interactive Program Sta-
tistics section, it is the cumulative CPU percentage of all transactions
through the listed program.
Cum Util (System) Cumulative CPU use (a running total).
Note: This is taken from the individual jobs and may differ slightly from
the total processing unit use on the workload page.
Cur Inl MPL (Transaction) The number of jobs waiting for an activity level (ineli-
gible) in the storage pool.
Cur MPL (Transaction) The number of jobs holding an activity level in the storage
pool.
DASD Ops/Sec (Component) Disk operations per second.
DASD Ops Per Sec Reads (Resource) Number of reads per second
DASD Ops Per Sec Writes (Resource) Number of writes per second
DB Fault (System, Component) Average number of database faults per second.
DB Pages (System, Component) Average number of database pages read per
second.
DB Read (Transaction) When listed in Physical I/O Counts column, it is the
number of database read requests while the job was in that state. When
listed in the Sync Disk I/O Rqs/Tns column, it is the average number of
synchronous database read requests per transaction.
DB Write (Transaction) When listed in the Sync Disk I/O Rqs/Tns column, it is the
average number of synchronous database write requests per trans-
action.
DB Wrt (Transaction) When listed in the Physical I/O Counts column, it is the
number of database write requests while the job was in that state. When
listed in the Synchronous Disk I/O Counts column, it is the number of
synchronous database write requests per transaction.
| DDM I/O (Component, Job Interval) The number of logical database I/O oper-
| ations for a distributed data management (DDM) server job.
DDM Svr Wait /Tns (Transaction) The average time, in seconds, that a source dis-
tributed data management (DDM) server job spent waiting for the target
system to respond to a request for data per transaction. This value

7-110 Performance Tools V4R2

 includes line time and time spent by the target system responding to the
request for data.
Dec (Transaction) The number of decimal overflow exceptions.
Decimal Data (Component) Data exception count per second. A data exception
occurs when data that is not valid is detected by arithmetic instructions.
Examples are signs or digit codes that are not valid in decimal
instructions, or an insufficient number of farthest left zeros in multiply
instructions.
Decimal Overflow (Component) Number of decimal overflows per second.
Description (Component) More detailed description of the exception type.
Detected Access Transmission Error (DTSE) In (Resource Interval) The number
of times the network termination 1 (NT1) end point notified the terminal
equipment (TE) of an error in data crossing the ISDN U interface from
the line transmission termination (LT) to the NT1 end point. The NT1
end point reports the errors to the TE through the maintenance channel
S1.
Detected Access Transmission Error (DTSE) Out (Resource Interval) The
number of times the network termination 1 (NT1) end point notified the
terminal equipment (TE) of an error in data crossing the ISDN U inter-
face from the NT1 end point to the LT. The NT1 end point reports the
errors to the TE through the maintenance channel S1.
Device (Component) Device identifier.
DIO/Sec Async (System) Number of asynchronous I/O operations per second.
DIO/Sec Sync (System) Number of synchronous I/O operations per second.
Disk Arm Seek Distance (Component) Average seek distance distributions per
hour:
0 Number of zero seeks
1/12 Number of seeks between 0 and 1/12 of the disk
1/6 Number of seeks between 1/12 and 1/6 of the disk
1/3 Number of seeks between 1/6 and 1/3 of the disk
2/3 Number of seeks between 1/3 and 2/3 of the disk
>2/3 Number of seeks greater than 2/3 of the disk
Disk Arms (System) The number of disk arms for this IOP.
Disk Capacity (Component) Average amount of disk space used or available.
MB Millions of bytes available on the disk.
Percent Percent of space available on the disk.
Disk Controllers (System) The number of disk storage controllers for this IOP.
| Disk CPU Util (System, Resource Interval) The percentage of CPU used by the
| disk unit.
Disk Feature (System) The type of disk (9332, 9335, and so on).
Disk I/O Async (System) Total number of asynchronous disk I/O operations.

Chapter 7. Performance Reports—Manager Feature 7-111

 Disk I/O per Second (System) Average number of physical disk I/O operations per
second.
Disk I/O Reads /Sec (Resource Interval) The average number of disk read oper-
ations per second by the disk IOP.
Disk I/O Requests (Transaction) The total number of synchronous and asynchro-
nous disk I/O requests issued by the jobs during the trace period.
Disk I/O Sync (System) Total number of synchronous disk I/O operations.
Disk I/O Writes /Sec (Resource Interval) The average number of disk write oper-
ations per second by the disk IOP.
Disk IOPs (System) The number of disk IOP controllers.
Disk mirroring (System) Indicates whether disk mirroring is active.
Disk Space Used (Resource Interval) The total disk space used in millions of bytes
for the entire system.
Disk transfer size (KB) (System) The average number of kilobytes transferred per
disk operation.
Disk utilization (System) The fraction of the time interval that the disk arms were
performing I/O operations.
Elapsed Seconds (Transaction, Component) The elapsed time in seconds. For the
Batch Job Analysis section of the Transaction Report, it is the number of
seconds elapsed from when the job started to when the job ended. For
the Concurrent Batch Job Statistics section of the Transaction Report, it
is the total elapsed time of all jobs in that job set.
Elapsed Time (Job Interval) The amount of time (minutes and seconds) for which
the job existed during the interval. This is the same as the interval
length unless the job started or ended during the interval, in which case
it is less.
Elapsed Time—Seconds (Transaction) Shows the time spent by the job, in the fol-
lowing columns:
Long Wait Elapsed times in the state (such as waiting for the next
transaction or lock-wait time).
Active/Rsp During transaction processing, the time the job spends
(either waiting or active) while it holds an activity level. At the
end of a transaction (on the transaction totals line), this is the
time the job spent processing the transaction in an activity
level, for long waits caused by locks, and in the ineligible
state.
Inel Wait The time the job spent in the ineligible wait state waiting for
an activity level.
EM3270 Wait /Tns (Transaction) The average, in seconds, of the time spent
waiting on the host system communications for Systems Network Archi-
tecture (SNA) and binary synchronous communications (BSC) 3270DE
per transaction. Program logic is required to determine if the emulation
program is communicating with the display or the host processing unit.
Because there are requirements on event-wait processing, not all transi-
tion combinations can be detected.

7-112 Performance Tools V4R2

 EORn (Transaction) Listed in the Wait Code column, End of response time for
transaction n. 1
EOTn (Transaction) Listed in the Wait Code column, End of transaction for
transaction for type n. 1
Estimated Exposr AP Not Jrnld (Component) System-estimated access path
recovery time exposure in minutes if no access paths were being jour-
naled by the system.
Estimated Exposr Curr System (Component) System-estimated access path
recovery time exposure in minutes.
Est Of AWS (Transaction) An estimate of the number of active work stations for
the trace period or interval. Any delay time greater than 600 seconds
has been rounded to 600 seconds. This technique is used to reduce the
effect of very casual users (those who may do intermittent work or leave
their work stations for long periods of time) on the estimate of active
work stations. This value is calculated as shown in Figure 7-80.

(AVGRSP + ACTIVE KEY/THINK)

AWS = TNS/HOUR x ─────────────────────────
36ðð

Figure 7-80. Equation for the Estimated Number of Active Work Stations

Event Wait /Tns (Transaction) The average time, in seconds, of the event-wait
time per transaction.
Often requests made by a job that runs on the system are made to
asynchronous jobs. These asynchronous jobs use an event to signal
completion of the request back to the requester. The event-wait time is
the time the requesting job waits for such a signal.
EVT (Transaction) Listed in the Wait Code column, Event Wait. This is a long
wait that occurs when waiting on a message queue.
Exception Type (Component) Type of program exception that results from the
internal microprogram instructions being run in internal microprogram
instructions procedure. Because these exceptions are monitored at a
low level within the system, it is difficult to associate these exceptions
with specific end-user operations. The counts are meaningful when the
processing unit time required to process them affects system perform-
ance. A variation in the counts may indicate a system change that could
affect performance. For example, a large variation in seize or lock
counts may indicate a job scheduling problem or indicate that contention
exists between an old application and a new one that uses the same
resources.
| Note: To see the seize and lock counts, you should collect the trace
| data by using the Start Performance Monitor (STRPFRMON)
| command and specifying TRACE(*ALL). Run the Print Trans-
| action Report (PRTTNSRPT) to list the objects and jobs that are
| holding the locks.

1 These codes are in the wait code column, but they are not wait codes. They indicate transaction boundary trace records. For more
information see Chapter 8, “Transaction Boundaries—Manager Feature” on page 8-1.

Chapter 7. Performance Reports—Manager Feature 7-113

 Exceptional wait (System) The average exceptional wait time, in seconds, per
transaction. An exceptional wait is that portion of internal response time
that cannot be attributed to the use of the processor and disk. An excep-
tional wait is caused by contention for internal resources of the system,
for example, waiting for a lock on a database record.
Note: This is a calculated value. If the sum of the constant and vari-
able wait time is greater than one second, you should run
STRPFRMON measurements with trace data collection and
compare the measured exceptional wait value, which
PRTTNSRPT provides, with this calculated value. If the values
are significantly different, use the value from PRTTNSRPT,
dividing it equally between constant and variable wait time.
Constant The portion of exceptional wait time held constant as
throughput increases.
Variable The portion of exceptional wait time that varies as
throughput increases.
Excp (Component, Transaction) For the Component Report, it is the total
number of program exceptions that occurred (see “Exception Occur-
rence Summary and Interval Counts” on page 7-27). For the Trans-
action Report, a Y in this column means that the transaction had
exceptions. The types of exceptions that are included are process
access group exceptions, and decimal, binary, and floating point over-
flow. See the transition report to see which exceptions the transaction
had.
Excp Wait (Transaction) The amount of exceptional wait time for the jobs in the job
set in seconds.
Excp Wait /Tns (Transaction) The average exceptional wait time, in seconds, per
transaction. This value is the sum of those waits listed under the Excep-
tional Wait Breakdown by Job Type part.
Excp Wait Sec (Transaction) The total amount of exceptional wait time in seconds
for the job.
Excs ACTM /Tns (Transaction) The average time, in seconds, of the excess
activity level time per transaction (for example, time spent in the active
state but not using the processing unit). If enough activity levels are
available and there is plenty of interactive work of higher priority to do, a
job waits longer for processing unit cycles. If the value is greater than .3,
look at jobs that correspond to particular applications for more informa-
tion. By looking at these jobs, you might be able to determine which
application’s jobs are contributing most to this value. Use the Trans-
action and Transition Reports for these jobs for additional information.
The formula for excessive activity-level time is shown in Figure 7-81.

| Active Time − [
| (multiplier X CPU X Beginning Activity Level) +
| (Number of synchronous disk I/O operations X .ð1ð)]
Figure 7-81. Formula for Excessive Activity-Level Time

| Note: If the beginning activity level is greater than 1, the multiplier

| equals 0.5. If the beginning activity level is any other value, the
| multiplier equals 1.

7-114 Performance Tools V4R2

 Expert Cache (System, Component) Directs the system to determine which objects
or portions of objects should remain in a shared main storage pool
based on the reference patterns of data within the object. Expert cache
uses a storage management tuner, which runs independently of the
system dynamic tuner, to examine overall paging characteristics and
history of the pool.
| Some values that you might see in this column are associated with the
| Work with Shared Pools (WRKSHRPOOL) command:
| Ÿ 0=*FIXED, which indicates the system does not dynamically adjust
| the paging characteristics of the storage pool. The system uses
| default values.
| Ÿ 3=*CALC, which indicates the system dynamically adjusts the
| paging characteristics of the storage pool for optimum performance.
Exposed AP System Journaled (Component) The number of exposed access
paths currently being journaled by the system.
Exposed AP System Not Journaled (Component) The number of exposed access
paths currently not being journaled by the system.
| /F (System, Resource Interval) The line speed of the protocol reported as
| full duplex. This indicator applies to the line speeds for an Ethernet
| (ELAN) token-ring (TRLAN) line, or an asynchronous transfer mode line.
Far End Code Violation (Resource Interval) The number of unintended code vio-
lations detected by the network termination 1 (NT1) end point for frames
transmitted to the NT1 end point on the interface for the T reference
point. The NT1 end point reports a violation to the termination equip-
ment (TE) through the maintenance channel S1.
File (Transaction) The file that contains the object.
Flp (Transaction) The number of floating point overflow exceptions.
Flp Overflow (Component) Number of floating point overflows per second.
Frame Retry (Resource Interval) The number of attempts to retransmit a frame to a
remote controller.
Frames Received Pct Err (Resource Interval) The percentage of frames received
in error. Errors can occur when the host system has an error or cannot
process received data fast enough.
Frames Received Total (Resource Interval) The total number of frames received
including frames with errors and frames that are not valid.
Frames Transmitted Pct Err (Resource Interval) The percentage of frames
retransmitted due to error.
Frames Transmitted Total (Resource Interval) The total number of frames trans-
mitted.
Functional Areas (System, Component, Transaction, Job Interval, Pool Interval)
For Report Selection Criteria, the list of functional areas selected to be
included (SLTFCNARA parameter) or excluded (OMTFCNARA param-
eter).
| /H (System, Resource Interval) The line speed of the protocol reported as
| half duplex. This indicator applies to the line speeds for an Ethernet
| (ELAN) token-ring (TRLAN) line, or an asynchronous transfer mode line.

Chapter 7. Performance Reports—Manager Feature 7-115

 HDW (Transaction) Listed in the Wait Code column, Hold Wait (job suspended
or system request). The job released a lock it had on the object named
on the next detail line of the report (OBJECT --). The job that was
waiting for the object is named on this line (WAITER --) along with the
amount of time the job spent waiting for the lock to be released.
High Srv Time (Resource Interval) The highest average service time in seconds for
a disk arm in the system.
High Srv Unit The disk arm with the highest service time.
High Util (Resource Interval) The percentage of use for the disk arm that has the
highest utilization.
High Util Unit (Component, Resource Interval) The disk arm with the highest utili-
zation.
High Utilization Disk (Component) Percent of utilization of the most utilized disk
arm during this interval.
High Utilization Unit (Component) Disk arm that had the most utilization during
this interval.
Holder Job Name (Transaction) The name of the job that held the object.
Holder Number (Transaction) The number of the job that held the object.
Holder Pool (Transaction) The pool that held the job while it was running.
Holder Pty (Transaction) The priority of the holder’s job.
Holder Type (Transaction) The type and subtype of the holder’s job.
Holder User Name (Transaction) The name of the user that held the object.
Holder’s Job Name (Lock) The name of the job holding the lock.
I Frames Recd per Sec (Resource Interval) The number of information frames
received per second.
I Frames Trnsmitd per Sec (Resource Interval) The number of information frames
transmitted per second.
I/O Wait (Resource Interval) The amount of time in which a given I/O request is
ready to be processed, but the disk arm is not yet available to perform
the request.
Incoming Calls Pct Retry (Resource Interval) The percentage of incoming calls
that were rejected by the network.
Incoming Calls Total (Resource Interval) The total number of incoming call
attempts.
Inel Time A-I/W-I (Transaction) The amount of time the job spent in the ineligible
state, either coming from time slice end (active-to-ineligible) or from the
wait state (wait-to-ineligible).
Inel Wait (Transaction) Listed in the Elapsed Time—Seconds column, the time the
job spent in the ineligible wait state waiting for an activity level.
Inter CPU Utilization (Component) Percentage of available processing unit time
used by the following types of jobs:
Ÿ Interactive
Ÿ Multiple requester terminal (MRT)

7-116 Performance Tools V4R2

 Ÿ System/36 environment interactive
Ÿ Pass-through
Ÿ Target distributed data management (DDM) servers
Ÿ Client Access servers
Note: For a multiple-processor system, this is the average use across
all processors.
IOP (Component) Input/output processor (IOP) Resource name and model
number for each communications IOP, DASD IOP, local workstation
IOP, and multifunction IOP. Communications IOP is the percent of CPU
used in the IOP. The percent does not necessarily mean that the IOP is
doing any data transfers. Some of the percent can be attributed to over-
head of an active line.
IOP Name/Line (Resource Interval) Input/output (IOP) processor resource name
and model number line.
IOP Name(Model) (Resource Interval) The input/output processor (IOP) identifica-
tion and the model number in parentheses.
IOP Name (System, Component) Input/Output processor (IOP) resource name.
IOP Name Network Interface (Resource Interval) The IOP name of the network
interface.
IOP Processor Util Comm (Component, Resource) Utilization of IOP due to com-
munications activity.
IOP Processor Util LWSC (Component, Resource) Utilization of IOP due to local
workstation activity.
IOP Processor Util DASD (Component, Resource) Utilization of IOP due to DASD
activity.
IOP Processor Util Total (Component, Resource Interval) The total percent of utili-
zation for each local workstation, disk, and communications IOP.
IOP Util (System) For the Disk Utilization section of the System Report, it is the
percentage of utilization for each input/output processor (IOP).
Note: For the multifunction I/O processors, this is utilization due to disk
activity only, not communications activity. For the System Model
Parameter section it is the fraction of the time interval the disk
IOP was performing I/O operations.
Itv End (Component, Transaction, Job Interval, Pool Interval, Resource Interval)
The time (hour and minute) when the data collection interval ended or a
vary off occurred. For the Exception Occurrence Summary and Interval
Counts of the Component Report, it is the ending time for the sample
interval in which STRPFRMON recorded the exception.
Job Maximum A-I (Pool Interval) The highest number of active-state to ineligible-
state transitions by a selected job in the pool or subsystem.
Job Maximum A-W (Pool) The highest number of active-to-wait state transitions by
a selected job in the pool or subsystem.
Job Maximum CPU Util (Pool Interval) The highest percentage of available proc-
essing unit time used by a selected job in the pool or subsystem.

Chapter 7. Performance Reports—Manager Feature 7-117

 Job Maximum Phy I/O (Pool Interval) The highest number of physical disk input
and output operations by a selected job in the pool or subsystem.
Job Maximum Rsp (Pool Interval) The highest response time in seconds per trans-
action by a selected job in the pool or subsystem. The response time is
the amount of time spent waiting for and using the resources divided by
the number of transactions.
Job Maximum Tns (Pool Interval) The highest number of transactions by a
selected job in the pool or subsystem.
Job Maximum W-I (Pool Interval) The highest number of wait-state to ineligible-
state transitions by a selected job in the pool or subsystem.
Job Name (Component, Transaction, Job Interval, Batch Job Trace) Name of the
job. In the Job Summary Report of the Transaction Report, a job (iden-
tical job name, user name, and job number) appears multiple times in
this list if the job uses the system Reroute Job (RRTJOB) command.
Job Number (Component, Transaction, Job Interval, Batch Job Trace) The number
of the job which the summary line describes. In the Transaction Report,
an asterisk (*) before the job number indicates the job signed on during
the measurement period. An asterisk (*) after the job number indicates
the job signed off during the measurement period.
Job Pty (Batch Job Trace) Priority of the job.
| Job Set (Transaction) The number of job sets is the number of batch jobs that
| could be active at any time during the trace period. If two jobs run
| sequentially, they show up as two jobs in the same job set. If two jobs
| run concurrently, they show up in two different job sets.
| Job Type (All Reports except where noted for the Transaction Report) Job type
and subtype.
Possible job type values include the following:
A Autostart
B Batch
| BD Batch immediate (Transaction only)
| Note: The batch immediate values are shown as BCI on the
| Work with Active Job display and as BATCHI on the
| Work with Subsystem Job display.
| BE Batch evoke (Transaction only)
| BJ Batch pre-start job (Transaction only)
C Programmable workstation application server, which includes
5250 emulation over APPC and Client Access host servers
running either APPC or TCP/IP
| A job is reported as a Client Access server if any of the fol-
| lowing items are true:
| Ÿ Incoming APPC evoke requests one of the server
| program names listed in the AS/400 Client Access Host
| Servers. This also applies to the pre-started jobs for the
| QSERVER, QCMN, and QSYSWRK subsystems that are
| already waiting for the named program.

7-118 Performance Tools V4R2

| Ÿ Incoming IP port number corresponds to one of the
| service name-description-port-numbers documented in
| AS/400 Client Access Host Servers. This also applies to
| the pre-started jobs for the QSERVER, QCMN, and
| QSYSWRK subsystems that are already waiting for the
| assigned IP port number.
| Ÿ Incoming IPX socket number corresponds to one of the
| service name-description-port-number documented in
| AS/400 Client Access Host Servers. This also applies to
| the pre-started jobs for the QSERVER, QCMN, and
| QSYSWRK subsystems that are already waiting for the
| assigned IPX port number.
| Ÿ Incoming 5250 display emulation jobs that come from
| APPC data streams sent by 5250 emulation under OS/2
| Communications Manager or WARP equivalent.
D Target distributed data management (DDM) server
| I Interactive. Interactive includes twinaxial data link control
| (TDLC), 5250 remote workstation, and 3270 remote work-
| station. For the Transaction Report, this includes twinaxial
| data link control (TDLC), 5250 remote workstation, 3270
| remote workstation, SNA pass-through, and 5250 TELNET.
L Licensed Internal Code task
M Subsystem monitor
| P SAN pass-through and 5250 TELNET pass-through. On the
| Transaction Report, these jobs appear as I (interactive).
R Spool reader
S System
| W Spool writer, which includes the spool write job, and if
| Advanced Function Printing (AFP) is specified, the print
| driver job.
WP Spool print driver (Transaction only)
X Start system job
Possible job subtype values include the following:
D Batch immediate job
E Evoke (communications batch)
J Pre-start job
P Print driver job
T Multiple requester terminal (MRT) (System/36 environment
only)
3 System/36
Noninteractive job types include:
Ÿ Autostart
Ÿ Batch

Chapter 7. Performance Reports—Manager Feature 7-119

 Ÿ Evoke
Ÿ Spool
Special interactive job categories include:
Ÿ Client Access server
Ÿ Distributed data management (DDM) server
Ÿ Interactive
Ÿ Multiple requester terminal (MRT)
Ÿ Pass-through
Ÿ System/36
Jobs (System, Component, Transaction, Pool Interval, Job Interval) The jobs
you specify. The format of the entries is jobnumber/username/jobname.
For the Report Selection Criteria report, it is the list of jobs selected to
be included (SLTJOB parameter) or excluded (OMTJOB parameter).
This does not include jobs selected by using the STLFCNARA or
OMTFCNARA parameter.
K per I/O (System, Resource Interval) The average number of kilobytes (1024
bytes) read or written for each disk I/O operation.
K/T /Tns Sec (Transaction) The average delay time, or time spent keying and
thinking between transactions for the job, in seconds. The value repres-
ents the time interval between active-to-wait and wait-to-active or wait-
to-ineligible job state transitions.
KB per I/O Read (Resource Interval) The average number of kilobytes (1 KB
equals 1024 bytes) transferred per read operation.
KB per I/O Write (Resource Interval) The average number of kilobytes (1024
bytes) transferred per write operation.
KBytes Transmitted IOP (Component, Resource Interval) Total kilobytes trans-
mitted from an IOP to the system across the bus.
KBytes Transmitted System (Component, Resource Interval) Total kilobytes
transmitted to the IOP from the system across the bus.
Key/Think (Transaction) The amount of time spent waiting for the work station user
by the program.
Key/Think /Tns (Transaction) The average think time and keying time (or the delay
time between transaction boundaries), in seconds, for the interactive
jobs.
L (Lock) Whether this is a lock or seize conflict. The column contains an L
if lock, blank if seize.
LAPD Pct Frames Recd in Error (Resource Interval) The percentage of frames
received in error (applies to D-channel only). Errors can occur when the
host system has an error or cannot process received data fast enough.
LAPD Pct Frames Trnsmitd Again (Resource Interval) The percentage of frames
retransmitted due to error (applies to D-channel only).
LAPD Total Frames Recd (Resource Interval) The total number of frames received
including frames with errors and frames that are not valid (applies to
D-channel only).

7-120 Performance Tools V4R2

LAPD Total Frames Trnsmitd (Resource Interval) The total number of frames
transmitted (applies to D-channel only).
Last 4 Programs in Invocation Stack (Transaction) The last four programs in the
program stack. For example, at the start of a transaction (such as when
the work station operator presses the Enter key), you see the program
names QT3REQIO, QWSGET, and the program that issued a read
operation. At the end of the transaction (such as when the program
writes to the display), you see QT3REQIO, QWSPUT, and the program
that wrote the display. See Appendix B, Defining Transaction Bounda-
ries, for more information about the transaction boundary.
Usually, the third or fourth program in the stack is the program shown in
the transaction summary PGMNAME data. However, if the Wait Code
column has a value, the program in the column labeled Last is the one
that caused the trace record.
If there is no program name in a column, the program name was the
same as the previous one in the column, and the name is omitted.
Length of Wait (Lock) The number of milliseconds the requestor waited for the
locked object.
Lgl I/O /Sec (Job Interval) The average number of logical disk I/O operations per-
formed per second by the job during the interval. This is calculated from
the logical disk I/O count divided by the elapsed time.
Library (Transaction) The library that contains the object.
Line Count (Job Interval) The number of lines printed by the selected
noninteractive jobs during the interval.
Line Descriptn (Resource Interval) Line description name.
Line Errors (Resource Interval) The total of all detected errors. Check the condition
of the line if this value increases greatly over time.
Line Speed (System, Resource Interval) The line speed in kilobits (1 kilobit = 1000
bits) per second.
Line Util (Resource Interval) The percent of available line capacity used by
transmit and receive operations.
LKRL (Transaction) Lock Released. The job released a lock it had on the
object named on the next detail line of the report (OBJECT --). The job
that was waiting for the object is named on this line (WAITER --) along
with the amount of time the job spent waiting for the lock to be released.
LKW (Transaction) Listed in the Wait Code column, Lock Wait. If there are a
number of these, or you see entries with a significant length of time in
the ACTIVE/RSP* column, additional analysis is necessary. The LKWT
report lines that precede this LKW report line show you what object is
being waited on, and who has the object.
LKWT (Transaction) Listed in the Wait Code column, Lock Conflict Wait. The
job is waiting on a lock conflict. The time (*/ time /*) is the duration of
the lock conflict and, though not equal to the LKW time, should be very
close to it. The holder of the lock is named at the right of the report line
(HOLDER --). The object being locked is named on the next report line
(OBJECT --).

Chapter 7. Performance Reports—Manager Feature 7-121

 Local End Code Violation (Resource Interval) The number of times an unintended
code violation was detected by the terminal equipment (TE) for frames
received at the interface for the ISDN S/T reference point.
Local Not Ready (Resource Interval) The percent of all receive-not-ready frames
that were transmitted by the host system. A large percentage often
means the host cannot process data fast enough (congestion).
Local work station IOP utilization The fraction of the time interval the work
station I/O processors are busy.
Local work station IOPs (System) The resource name and model number for
each local workstation IOP.
Lock Conflict (Component) Number of lock exceptions per second. Database
record contention is reflected in this count. For more information, run the
performance monitor with the TRACE(*ALL) option and use the
PRTTNSRPT and PRTLCKRPT commands.
This count could be very high, even under normal system operation.
Use the count as a monitor. If there are large variations or changes,
explore these variations in more detail.
Lock Wait /Tns (Transaction) The average time, in seconds, of the lock-wait time
per transaction. If the value is high, investigate with the transaction
detail calculation and the PRTLCKRPT command.
Logical (Job Interval) The number of logical disk I/O operations performed by
the selected interactive jobs during the interval.
Logical Database I/O Other (System) Other logical database operations per trans-
action. This includes operations such as update and delete.
Logical Database I/O Read (System) Logical database read operations per trans-
action.
Logical Database I/O Write (System) Logical database write operations per trans-
action.
Logical DB I/O (System) Average number of logical I/O operations per transaction.
Logical DB I/O Count (System) Number of database file input/output requests sent
by programs for each job type.
Note: Logical I/O contrasts with physical I/O shown elsewhere in these
reports. A logical I/O is a request sent at the program level that
might result in an access to auxiliary storage (DASD). A physical
I/O refers to those requests that result in access to auxiliary
storage.
Logical Disk I/O (Component) Number of logical disk operations (Get, Put, Update,
Other).
Logical I/O /Second (System) Average number of logical disk I/O operations per
second.
Logical I/O Per Second (Job Interval) The average number of logical disk I/O
operations performed per second by the selected noninteractive jobs
during the interval.

7-122 Performance Tools V4R2

Long Wait (Transaction) The time the job spent waiting for a system resource. An
example of a long wait would be a record-lock conflict. Also listed in the
Elapsed Time—Seconds column, it is the elapsed time in the state (such
as waiting for the next transaction or lock-wait time).
Long Wait Lck/Oth (Transaction) The amount of time the job spent waiting for a
system resource. An example of a long wait would be a record-lock con-
flict.
Loss of Frame Alignment (Resource Interval) The number of times a time period
equivalent to two 48-bit frames elapsed without detecting valid pairs of
line code violations.
MAC Errors (Resource Interval) The number of medium access control (MAC)
errors.
Main storage (MB) (System) The total main storage size, as measured in mega-
bytes (10242).
Max Util (System) Consistent use at or above the threshold value given will affect
system performance and cause longer response times or less
throughput. See the BEST/1 Capacity Planning Tool for a list of
threshold values.
Maximum (Transaction) The maximum value of the item that occurred in the
column.
Member (Transaction) The member that was involved in the conflict.
Minimum (Transaction) The minimum value of the item that occurred in the
column.
MRT Max Time (System) The time spent waiting, after MRTMAX is reached, by
jobs routed to a multiple requester terminal.
Note: No value appears in this column if job type is not MRT.
Nbr A-I (Transaction) The number of active-to-ineligible state transitions by the
job. This column shows the number of times that the job exceeded the
time-slice value assigned to the job, and had to wait for an activity-level
slot before the system could begin processing the transaction. If a value
appears in this column, check the work that the job was doing, and
determine if changes to the time-slice value are necessary.
Nbr Evt (Transaction) The number of event waits that occurred during the job
processing.
Nbr Jobs (Transaction) The number of jobs.
Nbr Sign offs (Transaction) The number of jobs that signed off during the interval.
Nbr Sign ons (Transaction) The number of jobs that signed on during the interval.
Nbr Tns (Transaction) The number of transactions in a given category.
Note: The values for transaction counts and other transaction-related
information shown on the reports you produce using the
PRTTNSRPT command may vary from the values shown on the
reports you produce using the PRTSYSRPT and PRTCPTRPT
commands. These differences are caused because the
PRTTNSRPT command uses trace data as input, while the
PRTSYSRPT and PRTCPTRPT commands use sample data as

Chapter 7. Performance Reports—Manager Feature 7-123

 input. See Appendix B, Defining Transaction Boundaries, for
additional information.
If there are significant differences in the values for transaction-related
information shown on these reports, do not use the data until you inves-
tigate why these differences exist.
Nbr W-I (Transaction) The number of wait-to-ineligible state transitions by the
job. This column shows how many times the job had to wait for a trans-
action.
NDB Read (Transaction) Listed in Physical I/O Counts column, it is the number of
nondatabase read requests while the job was in that state. Listed in the
Sync Disk I/O Rqs/Tns column, it is the average number of synchronous
nondatabase read requests per transaction.
NDB Write (Transaction) Listed in the Sync Disk I/O Rqs/Tns column, it is the
average number of synchronous nondatabase write requests per trans-
action.
NDB Wrt (Transaction) Listed in Physical I/O Counts column, the number of non-
database write requests while the job was in that state. Listed under
Synchronous Disk I/O Counts column, it is the number of synchronous
nondatabase write requests per transaction.
Non-DB Fault (System, Component) Average number of nondatabase faults per
second.
Non-DB Pages (System, Component) Average number of nondatabase pages read
per second.
Number (Transaction) The number of the job with which the transaction is asso-
ciated.
Number I/Os per Second (System) The number of I/Os per second for this partic-
ular IOP.
Number Jobs (Transaction) The number of batch jobs in the job set.
Number Lck Cft (Transaction) The number of lock-wait (including database record
lock) state conflicts that occurred during the job processing. If this
number is high, look at the Transaction and Transition Reports for the
job to see how long the lock-wait state conflicts were lasting. In addition,
you can do further investigation using the reports produced when you
use the PRTLCKRPT command.
Number Lck Conflict (Transaction) The number of times the job had a lock con-
flict.
Number Locks (Transaction) The number of locks attributed to interactive or
noninteractive waiters.
Number of batch jobs (System) The average number of active batch jobs. A batch
job is considered active if it averages at least one I/O per 5 minutes.
Number of Jobs (System) Number of jobs.
Number Seizes (Transaction) The number of seizes attributed to interactive or
noninteractive waiters.
Number Sze Cft (Transaction) The number of seize/lock conflicts that occurred
during the job processing. If this number is high, look at the Transaction
and Transition Reports for the job to see how long the conflicts lasted,

7-124 Performance Tools V4R2

 the qualified name of the job that held the object, the name and type of
object being held, and what the job was waiting for.
Number Sze Conflict (Transaction) The number of times the job had a seize con-
flict.
Number Tns (System, Transaction) Total number of transactions processed. For
example, in the System Report it is the total number of transactions
processed by jobs in this pool. In the Transaction Report it is the
number of transactions associated with the program.
Number Traces (Batch Job Trace) Number of traces.
Number Transactions (System) Total number of transactions processed.
Object File (Transaction) The file that contains the object.
Object Library (Transaction) The library that contains the object.
Object Member (Transaction) The member that was involved in the conflict.
Object Name (Lock) The name of the locked object.
Object RRN (Transaction) The relative record number of the record involved in the
conflict.
Object Type (Transaction, Lock) The type of the locked object. The following are
possible object types:
AG Access group
CB Commit block
CBLK Commit block
CD Controller description
CLS Class
CMD Command
CTLD Controller description
CTX Context
CUD Control unit description
CUR Cursor
DEVD Device description
DS Data space
DSI Data space index
DTAARA Data area
EDTD Edit description
FILE File
JOBD Job description
JOBQ Job queue
JP Journal port
JRN Journal
JRNRCV Journal receiver

Chapter 7. Performance Reports—Manager Feature 7-125

 JS Journal space
LIB Library
LIND Line description
LUD Logical unit description
MBR Member
MEM Database file member
MSGF Message file
MSGQ Message queue
ND Network description
OCUR Database operational cursor
OUTQ Output queue
PGM Program
PROG Program
PRTIMG Print image
QDAG Composite piece - access group
QDDS Composite piece - data space
QDDSI Composite piece - dta spe index
QTAG Temporary - access group
QTDS Temporary - data space
QTDSI Temporary - data space index
SBSD Subsystem description
TBL Table
Omit Parameters (System, Component, Transaction, Job Interval, Pool Interval)
The criteria used to choose the data records to be excluded from the
report. The criteria are generally specified using an OMTxxx parameter
of the command. Only nondefault values (something other than *NONE)
are printed. If a parameter was not specified, it does not appear on the
report.
Op per Second (System) Average number of disk operations per second.
Other Wait /Tns (Transaction) The average time, in seconds, spent waiting that
was not in any of the previous categories per transaction. For example,
the time spent waiting during a save/restore operation when the system
requested new media (tape or diskette).
Outgoing Calls Pct Retry (Resource Interval) The percentage of outgoing calls
that were rejected by the network.
Outgoing Calls Total (Resource Interval) The total number of outgoing call
attempts.
Over commitment ratio (System) The main storage over commitment ratio (OCR).
PAG (Transaction) The number of process access group faults.

7-126 Performance Tools V4R2

PAG Fault (Component, Job Interval) In the Exception Occurrence Summary of the
Component Report, it is the total number of times the program access
group (PAG) was referred to, but was not in main storage. In the Excep-
tion Occurrence Summary of the Component Report, it is the number of
faults involving the process access group per second.
Page Count (Job Interval) The number of pages printed by the selected
noninteractive jobs during the interval.
Pct CPU By Categories (Transaction) The percentage of available processing unit
time used by the transactions that fell into the various categories. See
the ANALYSIS by Interactive Transaction Categories part of the System
Summary Data Section for an explanation of the categories.
Pct Data Characters Received in Error (Resource Interval) The percent of data
characters received with error.
Pct Data Characters Transmitted in Error (Resource Interval) The percent of
data characters transmitted with error.
Pct Ex-Wt /Rsp (Transaction) The percentage of the response time that is due to
exceptional wait.
Pct Of Tns Categories (Transaction) The percentage of all transactions that fell
into the various categories. See the Analysis by Interactive Transaction
Categories part of the System Summary Data Section for an explanation
of the categories.
Pct PDUs Received in Error (Resource Interval) The percent of protocol data units
(PDUs) received in error during the time interval. These errors can occur
if the host system has errors or cannot receive data fast enough (con-
gestion).
Note: A protocol data unit (PDU) for asynchronous communications is
a variable-length unit of data that is ended by a protocol control
character or by the size of the buffer.
Pct Poll Retry Time (Resource Interval) The percent of the time interval the line
was unavailable while the IOP waited for a response from a work station
controller (or remote AS/400 system) that was in disconnect mode.
Note: To minimize this lost time:
Ÿ Vary on only the controllers that are turned on.
Ÿ Turn on all controllers.
Ÿ Use the Change Line Description (SDLC) (CHGLINSDLC) command
to set the connect poll timer to a small value (reduces wait time).
Ÿ Use the Change Controller Description (CHGCTLxxxx) command
(where xxxx is APPC, FNC, RWS, or RTL, as appropriate) to set the
NDMPOLLTMR value to a large value (increases time between
polls).
Pct Tns (Transaction) The percentage of the total transactions. For the System
Summary section of the Job Summary Report, the transactions are
within the given trace period with the given purge attribute. For the Inter-
active Program Transaction Statistics section of the Job Summary
Report, the percentage of transactions that were associated with a
program. For the Job Statistics section, it is the percentage of total

Chapter 7. Performance Reports—Manager Feature 7-127

 transactions that were due to this job. For the Interactive Program Sta-
tistics section, it is all transactions that were associated to a program.
Percent Errored Seconds (Resource Interval) The percentage of seconds in which
at least one Detected Access Transmission (DTSE) in or out error
occurred.
Percent Frames Received in Error (Resource Interval) The percent of all received
frames that were received in error. Errors can occur when the host
system has an error or cannot process received data fast enough (con-
gestion).
Percent Full (System) Percentage of disk space capacity in use.
Percent I Frames Trnsmitd in Error (Resource Interval) The percent of trans-
mitted information frames that required retransmission. Retransmissions
can occur when a remote device has an error or cannot process
received data fast enough (congestion).
Percent Severely Errored Seconds (Resource Interval) The percent of seconds in
which at least three Detected Access Transmission (DTSE) in or out
errors occurred.
Percent transactions (dynamic no) (System) A measure of system main storage
utilization. The percent of all interactive transactions that were done with
the purge attribute of dynamic NO.
Percent transactions (purge no) (System) A measure of system main storage
utilization. The percent of all interactive transactions that were done with
the purge attribute of NO.
Percent transactions (purge yes) (System) A measure of system main storage
utilization. The percent of all interactive transactions that were done with
the purge attribute of YES.
Percent Util (System) Average disk arm utilization (busy). Consistent use at or
above the threshold value provided for disk arm utilization affects
system performance, which causes longer response times or less
throughput. See utilization guidelines and thresholds in the BEST/1
Capacity Planning Tool book for a list of threshold values.
Note: The percent busy value is calculated from data measured in the
I/O processor. When comparing this value with percent busy
reported by the Work with Disk Status (WRKDSKSTS)
command, some differences may exist. The WRKDSKSTS
command estimates percent busy based on the number of I/O
requests, amount of data transferred, and type of disk unit.
The system-wide average utilization does not include data for mirrored
arms in measurement intervals for which such intervals are either in
resuming or suspended status.
Perm Write (Component, Job Interval) The number of permanent write operations
performed for the selected jobs during the interval.
Permanent writes per transaction (System) The average number of permanent
write operations per interactive transaction.
Physical I/O Count (Transaction, Batch Job Trace) For the Job Summary section
of the Batch Job Trace Report, the number of synchronous and asyn-
chronous disk operations (reads and writes). For the Transition Report,

7-128 Performance Tools V4R2

 the next five columns provide information about the number of synchro-
nous and asynchronous disk I/O requests while the job was in the given
state. The first line is the synchronous disk I/O requests, and the second
line is the asynchronous disk I/O requests.
DB Read The number of database read requests while the job was in
that state.
DB Wrt The number of database write requests while the job was in
that state.
NDB Read The number of nondatabase read requests while the job was
in that state.
NDB Wrt The number of nondatabase write requests while the job was
in that state.
Tot The total number of DB Read, DB Wrt, NDB Read, and NDB
Wrt requests.
Pl (Component, Transaction, Job Interval, Pool Interval) The number of the
pool in which the subsystem or job ran.
Pool (Transaction, Job Interval, Batch Job Trace) The number of the pool
containing the transaction (for example, in which the job ran.)
Pool ID (System) Pool identifier.
Pool ID Faults (Component) User pool that had the highest page fault rate.
Pool Mch Faults/Sec (Component) Average number of machine pool page faults
per second.
Pool size (KB) (System, Component) For the Storage Pool Activity section of the
Component Report it is the initial pool size in kilobytes (1024 bytes). For
the System Model Parameters section of the System Report, it is the
total size in kilobytes of all pools that incurred interactive job activity.
Pool User Faults/Sec (Component) Average number of user pool page faults per
second, for the user pool with highest fault rate during this interval.
Pools (System, Component, Transaction, Job Interval, Pool Interval) In the
Report-Selection Criteria section, the list of pools selected to be included
(SLTPOOLS parameter) or excluded (OMTPOOLS parameter). Other-
wise, the pools you specify. The values can be from 1 through 16.
Prg (Transaction) The purge attribute of the jobs.
Printer Lines (System, Job Interval) The number of lines printed by the job during
the interval.
Printer Pages (System, Job Interval) The number of pages printed by the job
during the interval.
Priority (System, Transaction) The priority of the job.
Program (Transaction) The name of the program with which the transaction is
associated.
Program Name (Transaction) For the Job Summary section of the Transaction
Report, the name of the program in control at the start of the trans-
action. Other programs may be used during the transaction. For the
Transaction Report section, the name of the program active at the start
of the transaction. If ADR=UNKNWN (address unknown) is shown under

Chapter 7. Performance Reports—Manager Feature 7-129

 the column, the program was deleted before the trace data was dumped
to the database file. If ADR=000000 is shown under the column, there
was not enough trace data to determine the program name, or there
was no program active at that level in the job when the trace record was
created.
Protocol (System) Line protocol.
Ÿ SDLC
Ÿ ASYNC
Ÿ BSC
Ÿ X25
Ÿ TRLAN
Ÿ ELAN (Ethernet)
Ÿ IDLC
Ÿ DDI
Ÿ FRLY
Pty (Component, Transaction, Job Interval) Priority of the job. For the Con-
current Batch Job Statistics section of the Transaction Report, it is the
priority of the jobs in the job set.
Purge (Transaction) The purge attribute of the jobs.
PWrt (Transaction) The number of permanent write I/O operations.
Queue Length (Resource Interval) The average number of I/O requests that had to
wait in the queue for this unit.
Rank (Transaction) The order. For the Job Summary section, it is the order of
the program according to the number of transactions. For the Job Statis-
tics section, it is the order of the job. For the Interactive Program Statis-
tics section, it is the order of the program. For the Individual Transaction
Statistics section, it is the order of the transaction according to the data
being put in order by importance. For the Largest Seize/Lock Conflicts
section, it is the order of the seize or lock conflict.
Ratio of write disk I/O to total disk I/O (System) The fraction of the total disk
activity that is due to writing data to the disks.
Reads per Second (Resource Interval) The average number of disk read oper-
ations performed per second by the disk arm.
Receive CRC Errors (Resource Interval) The number of received frames that con-
tained a cycle redundancy check (CRC) error. This indicates that the
data was not received error free.
Record Number (Lock) For database file members, the relative record number of
the record within the database file member.
Remote LAN Pct Frames Recd (Resource Interval) The number of frames
received from a local area network (LAN) connected to the locally
attached LAN.
Remote LAN Pct Frames Trnsmitd (Resource Interval) The number of frames
transmitted to a local area network (LAN) connected to the locally
attached LAN.

7-130 Performance Tools V4R2

Remote Not Ready (Resource Interval) The percentage of all receive-not-ready
frames that were received by the host system. A large percentage often
means the remote device cannot process data fast enough (congestion).
Remote Seq Error (Resource Interval) The percent of frames received out of order
by a remote device or system. This can occur when the remote device
or system cannot process data fast enough.
Requestor’s Job Name (Lock) The name of the job requesting the locked object
(the same as in the detail listing).
Reset Packets Recd (Resource Interval) The number of reset packets received by
the network. Reset packets are packets retransmitted because an error
occurred.
Reset Packets Trnsmitd (Resource Interval) The number of reset packets trans-
mitted by the network.
Response (System) Average system response (service) time.
Response Sec Avg and Max (Transaction) The average (AVG) and maximum
(MAX) transaction response time, in seconds, for the job. The average
response time is calculated as the sum of the time between each pair of
wait-to-active and active-to-wait transitions divided by the number of
pairs that were encountered for the job. The MAX response time is the
largest response time in the job.
Response Seconds (System) Average response time in seconds per transaction.
Rsp (Component) Average interactive transaction response time in seconds.
Rsp Time (Component, Resource Interval) The average external response time (in
seconds). For the Local Work Station IOP Utilizations section of the
Resource Interval Report, it is the response time for work stations on
this controller. For the Remote Work Stations section of the Component
Report, it is the response time for this work station.
Rsp Timer Ended (Resource Interval) The number of times the response timer
ended waiting for a response from a remote device.
Rsp/Tns (Component, Transaction, Job Interval) The average response time
(seconds) per transaction. For the Job Summary section of the Job
Interval Report, it is the response time per transaction for the selected
interactive jobs during the interval (the amount of time spent waiting for
or using the system resources divided by the number of transactions
processed). This number will not be accurate unless at least several
seconds were spent processing transactions.
S/L (Transaction) Whether the conflict was a seize (S) or lock (L) conflict.
Seize and Lock Conflicts (Batch Job Trace) Number of seize conflicts and lock
waits.
Seize Conflict (Component) Number of seize exceptions per second. For more
detailed information, you can run the performance monitor with the
TRACE(*ALL) option, and use the PRTTNSRPT or PRTLCKRPT com-
mands.
This count could be very high, even under normal system operation.
Use the count as a monitor. If there are large variations or changes,
explore these variations in more detail.

Chapter 7. Performance Reports—Manager Feature 7-131

 Seize Hold Time (Transaction) The amount of time that the transaction held up
other jobs in the system by a seize or lock on an object.
Seize Wait /Tns (Transaction) The average time, in seconds, for all seize-lock con-
flicts that occur during an average transaction. More than one seize-lock
conflict can occur during a single transaction for the same job. If this
number is high, investigate those jobs with seize conflicts. The Trans-
action Report lists each conflict that occurs, the name of the holder, and
the name of the object held.
For the Transaction by 5-Minute Intervals section of the Job Summary
Report, it is the average seize wait time per transaction in seconds. This
is the average amount of time that the transactions spent in a seize/lock
conflict. If this number is high, look at the Transaction and Transition
Reports for the jobs that are causing the excessive wait time.
Select Parameters (System, Component, Transaction, Job Interval, Pool Interval)
The criteria used to choose the data records to be included in the report.
The criteria are generally specified using an SLTxxx parameter of the
command. Only nondefault values (something other than *ALL) are
printed. If a parameter is not specified, it does not appear on the report.
Sequence Error (Resource Interval) The number of frames received that contained
sequence numbers indicating that frames were lost.
Short Frame Errors (Resource Interval) The number of short frames received. A
short frame is a frame that has fewer octets between its start flag and
end flag than are permitted.
Short Wait /Tns (Transaction) The average time, in seconds, of short (active) wait
time per transaction.
For the Interactive Program Statistics section, if the value is high, it may
be due to the use of data queues or to the use of DFRWRT(*NO) or
RSTDSP(*YES) in the program display files.
Short WaitX /Tns (Short wait extended) (Transaction) The average time, in
seconds, of wait time per transaction that resulted due to a short (active)
wait that exceeded 2 seconds, and caused a long wait transition to
occur. The activity level has been released but this time is still counted
against your total response time. Waits on data queues or the use of
DFRWRT(*NO) and/or RSTDSP(*YES) in the display files could be
reasons for this value to be high.
Size (Component) Decimal data overflow and underflow exceptions per
second. An indication of improper field size on numeric calculations.
Size (K) (System, Pool Interval) The size of the pool in kilobytes (1024 bytes).
Size (M) (System) Disk space capacity in millions of bytes.
SMAPP ReTune (Component) System-managed access path protection tuning
adjustments.
SOTn (Transaction) Listed in the Wait Code column, Start of transaction n. 2

2 These codes are in the wait code column, but they are not wait codes. They indicate transaction boundary trace records. For more
information see Chapter 8, “Transaction Boundaries—Manager Feature” on page 8-1.

7-132 Performance Tools V4R2

Spool CPU seconds per I/O (System) The average number of system processing
unit seconds used by all spool jobs for each I/O performed by a spool
job.
Spool database reads per second (System) The average number of read oper-
ations to database files per second of spool processing.
Spool I/O per second (System) The average number of physical disk I/O oper-
ations per second of spool processing.
Srv Time (Component) Average disk service time per request in seconds not
including the disk wait time. See Figure 7-9 on page 7-23 for disk
response time.
Start (Transaction) The time the job started.
Started (Transaction) The time of the first record in the trace data, in the form
HH.MM.SS (hours, minutes, seconds).
State (Transaction) The three possible job states are shown in Figure 7-82.

┌─────5 W (wait state) not holding an activity level

│
│ ┌───5 A (Active or wait state) holding an activity
│ │ level
│ │ ┌──5 I (ineligible state) waiting for an
│ │ │ activity level
│ │ │ The type of transition is shown below.
│ │ │ The direction and position of the arrow
│ │ │ indicates which transition was made.
│ │ │
│ │ │
W A I
────
Figure 7-82. Possible Job States

Figure 7-83 shows the possible job state transitions. For example, from
W to A is y, or yes, which means it is possible for a job to change from
the wait state to the active state.

To state
A = Active state
A W I W = Wait state
From I = Ineligible state
state A y y y

W y - y

I y - -
RV2S087-0

Figure 7-83. AS/400 Job State Transitions

State Transitions A-A (Batch Job Trace) Number of active-to-active transitions.

State Transitions A-I (Batch Job Trace) Number of active-to-ineligible transitions.
Stop (Transaction) The time the job ended.
Stopped (Transaction) The time of the last record in the trace data, in the form
HH.MM.SS (hours, minutes, seconds).
Subsystem Name (Pool Interval) The name of the subsystem.

Chapter 7. Performance Reports—Manager Feature 7-133

 Subsystems (System, Component, Pool Interval) For the System Report, the sub-
system names you specify. Each name is a 10-character name. For the
Component Report, the list of subsystems selected to be included
(SLTSBS parameter) or excluded (OMTSBS parameter).
Sum (Transaction) Listed in the Sync Disk I/O Rqs/Tns column, the sum of
the averages of the synchronous DB READ, DB WRITE, NDB READ,
and NDB WRITE requests (the average number of synchronous I/O
requests per transaction for the job).
SWX (Transaction) Listed in the Wait Code column, Short Wait Extended. The
short wait has exceeded a 2-second limit and the system has put the
transaction into a long wait. This long wait must be charged to the trans-
action response time. In most cases, this active-to-wait transaction does
not reflect a transaction boundary.
Sync (Job Interval) The number of synchronous disk I/O operations performed
by the selected interactive jobs during the interval.
Sync DIO /Tns (Transaction) The average number of synchronous I/O requests per
transaction during the interval.
Sync Disk I/O (System, Component, Transaction) Synchronous disk I/O operations.
Sync Disk I/O per Second (Component) Average synchronous disk I/O operations
per second.
Sync Disk I/O Requests (Transaction) The total number of synchronous disk I/O
requests for the given combination of priority, job type, and pool.
Sync Disk I/O Rqs/Tns (Transaction) The next five columns provide information
about the number of synchronous disk I/O requests per transaction:
DB Read The average number of synchronous database read requests
per transaction.
DB Write The average number of synchronous database write requests
per transaction.
NDB Read The average number of synchronous nondatabase read
requests per transaction.
NDB Write The average number of synchronous nondatabase write
requests per transaction.
Sum The sum of the averages of the synchronous DB READ, DB
WRITE, NDB READ, and NDB WRITE requests (the average
number of synchronous I/O requests per transaction for the
job).
Sync I/O /Elp Sec (Transaction) The average number of synchronous disk I/O
requests for all jobs, per second of elapsed time used by the jobs.
Sync I/O /Sec (Job Interval) The average number of synchronous disk I/O oper-
ations performed per second by the job during the interval. This is calcu-
lated from the synchronous disk I/O count divided by the elapsed time.
Sync I/O Per Second (Job Interval) The average number of synchronous disk I/O
operations performed per second by the selected noninteractive jobs
during the interval.

7-134 Performance Tools V4R2

Synchronous DBR (System, Transaction, Job Interval, Pool Interval) The average
number of synchronous database read operations. It is the total synchro-
nous database reads divided by the total transactions. For the Pool
Interval and Job Interval Reports, it is calculated per transaction for the
job during the intervals. For the System Report, it is calculated per
second. For the Transaction (Job Summary) it is calculated per trans-
action. Listed under Average DIO/Transaction, the average number of
synchronous database read requests per transaction. This field is not
printed if the jobs in the system did not process any transactions.
Synchronous DBW (System, Transaction, Job Interval, Pool Interval) The average
number of synchronous database write operations. It is the total syn-
chronous database writes divided by the total transactions. For the Pool
Interval and Job Interval Reports, it is calculated per transaction for the
job during the intervals. For the System Report, it is calculated per
second. For the Transaction (Job Summary) it is calculated per trans-
action. Listed under Average DIO/Transaction, the average number of
synchronous database read requests per transaction. This field is not
printed if the jobs in the system did not process any transactions.
Synchronous DIO / Act Sec (System, Transaction) The number of synchronous
disk I/O operations per active second. The active time is the elapsed
time minus the wait times.
Synchronous DIO / Ded Sec (Transaction) The estimated number of synchronous
disk I/O operations per second as if the job were running in dedicated
mode. Dedicated mode means that no other job would be active or in
contention for resources in the system.
Synchronous DIO / Elp Sec (Transaction) The number of synchronous disk I/O
operations per elapsed second.
Synchronous Disk I/O Counts (Transaction) The next five columns provide infor-
mation about the number of synchronous disk I/O requests per
transaction:
DB Read The number of synchronous database read requests per
transaction.
DB Wrt The number of synchronous database write requests per
transaction.
NDB Read The number of synchronous nondatabase read requests per
transaction.
NDB Wrt The number of synchronous nondatabase write requests per
transaction.
Sum The sum of the synchronous DB Read, DB Wrt, NDB Read,
and NDB Wrt requests (the number of synchronous I/O
requests per transaction).
Synchronous disk I/O per transaction (System, Transaction) The average
number of synchronous physical disk I/O operations per interactive
transaction.
Synchronous Max (Transaction) The maximum number of synchronous DBR,
NDBR, and WRT I/O requests encountered for any single transaction by
that job. If the job is not an interactive or autostart job type, the total disk
I/O for the job is listed here.

Chapter 7. Performance Reports—Manager Feature 7-135

 Synchronous NDBR (System, Transaction, Job Interval, Pool Interval) The
average number of synchronous nondatabase read operations per trans-
action for the jobs in the system during the interval. For the Transaction
Report, the operations on the disk per transaction for the selected jobs
in the pool. This is calculated from the synchronous nondatabase read
count divided by the transactions processed. This field is not printed if
the jobs in the system did not process any transactions.
Synchronous NDBW (System, Job Interval, Pool Interval) The average number of
synchronous nondatabase write operations on the disk per transaction
for the selected jobs in the pool. For the System Report, it is the oper-
ations per transaction for the jobs in the system during the interval. This
is calculated from the synchronous nondatabase write count divided by
the transactions processed. This field is not printed if the jobs in the
system did not process any transactions.
Synchronous Sum (Transaction) The sum of the averages of the synchronous
DBR, NDBR, and WRT requests (the average number of synchronous
I/O requests per transaction for the job).
Synchronous wrt (Transaction) The average number of synchronous database
and nondatabase write requests per transaction.
System CPU per transaction (seconds) (System) The average number of system
processing unit seconds per interactive transaction.
System disk I/O per transaction (System) The total number of physical disk I/O
operations attributed to the system per interactive transaction.
System Starts (Component) The number of start journal operations initiated by the
system.
System Stops (Component) The number of stop journal operations initiated by the
system.
System Total (Component) The total number of journal deposits resulting from
| system-journaled objects. These are the deposits performed by system-
| managed access path protection (SMAPP).
System ToUser (Component) The number of journal deposits resulting from
system-journaled objects to user-created journals.
SZRL (Transaction) Listed in the Wait Code column, Seize/Lock Released.
The job released a lock it had on the object named on the next detail
line of the report (OBJECT --). The job that was waiting for the object is
named on this line (WAITER --) along with the amount of time the job
spent waiting for the seize conflict to be released.
SZWT (Transaction) Listed in the Wait Code column, Seize/Lock Conflict Wait.
The job is waiting on a seize/lock conflict. The time (*/ time /*) is the
duration of the seize/lock conflict, and is included in the active time that
follows it on the report. The holder of the lock is named at the right of
the report line (HOLDER --). The object being held is named on the
next report line (OBJECT --).
| Thread (Job Summary, Transaction, Transition) A thread is a unique flow of
| control within a process. Every job has an initial thread associated with
| it. Each job can start one or more secondary threads. The user name
| column changes to User name / Thread if a secondary thread is avail-
| able.

7-136 Performance Tools V4R2

| The system assigns the thread number to a job as follows:
| Ÿ The system assigns thread IDs sequentially. When a job is started
| that uses a job structure that was previously active, the thread ID
| that is assigned to the initial thread is the next number in the
| sequence.
| Ÿ The first thread of a job is assigned a number.
| Ÿ Any additional threads from the same job are assigned a number
| that is incremented by 1. For example:
| Job Name User Name/ Job Number
| Thread
| QJVACMDSRV SMITH ð23416
| QJVACMDSRV ððððððð6 ð23416
| QJVACMDSRV ððððððð7 ð23416
| QJVACMDSRV ððððððð8 ð23416
| A thread value greater than 1 does not necessarily mean the job
| has had that many threads active at the same time. To determine
| how many threads are currently active for the same job, use the
| WRKACTJOB, WRKSBSJOB, or WRKUSERJOB commands to find
| the multiple three-part identifiers with the same job name.
Time (Transaction) The time when the transaction completed, or when a seize
or lock conflict occurred. Also, a column heading that shows the time the
transition from one state to another occurred, in the HH.MM.SS.mmm
arrangement.
Tns (Component, Pool Interval) The total number of transactions processed
by the selected jobs in the pool or subsystem.
Tns Count (Component, Job Interval) The number of transactions performed by the
selected interactive jobs during the interval.
Tns/Hour (Component, Transaction, Job Interval) The average number of trans-
actions per hour processed by the selected interactive jobs during the
interval.
Tns/Hour Rate (System) Average number of transactions per hour.
TOD of Wait (Lock) The time of day of the start of the conflict.
Tot (Transaction) Listed in Physical I/O Counts column, the total number of
DB Read, DB Wrt, NDB Read, and NDB Wrt requests.
Tot Nbr Tns (Transaction) The total number of transactions the PRTTNSRPT
program determined from the input data that were accomplished for the
job.
Total (Component) Total exception counts for the reporting period.
Total /Job (Transaction) The total (sum) of the items in the column for the job.
Total characters per transaction (System) The average number of characters
either read from or written to display station screens per interactive
transaction.
Total CPU Sec /Sync DIO (Transaction) The ratio of total CPU seconds divided by
the total synchronous disk I/O requests.

Chapter 7. Performance Reports—Manager Feature 7-137

 Total CPU Utilization (System, Component) Percentage of available processing
unit time used by interactive jobs, batch jobs, all system jobs, and
Licensed Internal Code tasks. For a multiple-processor system this is
the average use across all processors.
For a multiple-processor system, Total CPU Utilization is replaced by a
utilization value for each processor in the system. Here is an example of
this part of the display for a system with two processors:
Average CPU utilization . . . . . : 41.9
CPU 1 utilization . . . . . . . . : 41.7
CPU 2 utilization . . . . . . . . : 42.2
Note: This value is taken from a system counter. Other processing unit
uses are taken from the individual job work control blocks
(WCBs). These totals may differ slightly.
Total Data Characters Received (Resource Interval) The number of data charac-
ters received successfully.
Total Data Characters Transmitted (Resource Interval) The number of data char-
acters transmitted successfully.
Total fields per transaction (System) The average number of display station fields
either read from or written to per interactive transaction.
Total Frames Recd (Resource Interval) The number of frames received, including
frames with errors and frames that are not valid.
Total I Frames Trnsmitd (Resource Interval) The total number of information
frames transmitted.
Total I/O (System) Sum of the read and write operations.
Total PDUs Received (Resource Interval) The number of protocol data units
(PDUs) received during the time interval.
Note: A protocol data unit (PDU) for asynchronous communications is
a variable-length unit of data that is ended by a protocol control
character or by the size of the buffer.
Total Physical I/O per Second (Resource Interval) The average number of phys-
ical disk I/O operations performed per second by the disk arm.
Total Responses (Component, Resource Interval) The total number of transactions
counted along with the average response time for all active work
stations or devices on this controller for the report period.
Total Seize/Wait Time (Component) The response time in milliseconds for each
job.
Total Tns (Component) Number of transactions processed in this pool.
Transaction Response Time (Sec/Tns) (Transaction) The response time in
seconds for each transaction. This value includes no communications
line time. Response times measured at the work station exceed this time
by the data transmission time (the time required to transmit data from
the work station to the processing unit and to transmit the response data
back to the work station from the processing unit).
Transactions per hour (local) (System) The interactive transactions per hour attri-
buted to local display stations.

7-138 Performance Tools V4R2

 Transactions per hour (remote) (System) The interactive transactions per hour
attributed to remote display stations.
Transmit/Receive/Average Line Util (Resource Interval) In duplex mode, the per-
centage of transmit line capacity used, the percentage of receive line
capacity used, and the average of the transmit and receive capacities.
TSE (Transaction) Listed in the Wait Code column, Time Slice End. The
program shown in the stack entry labeled LAST is the program that went
to time slice end.
| Typ (Component, Transaction) The system job type and subtype. The Com-
| ponent Report allows only one character in this column. The Transaction
| Report allows two characters. The Transaction Report reports the job
| type and job subtype directly from the QAPMJOBS fields. The Compo-
| nent Report takes the job type and job subtype values and converts it to
| a character that may or may not be the value from the QAPMJOBS
| field. Table 7-1 on page 7-16 shows the following information in table
| format. The possible job types are:
A Autostart
B Batch
| BD Batch immediate (Transaction only)
| Note: The batch immediate values are shown as BCI on the
| Work with Active Job display and as BATCHI on the
| Work with Subsystem Job display.
| BE Batch evoke (Transaction only)
| BJ Batch pre-start job (Transaction only)
C Programmable work station application server, which includes
5250 emulation over APPC and Client Access host servers
running either APPC or TCP/IP
| A job is reported as a Client Access server if any of the fol-
| lowing items are true:
| Ÿ Incoming APPC evoke requests one of the server
| program names listed in the AS/400 Client Access Host
| Servers. This also applies to the pre-started jobs for the
| QSERVER, QCMN, and QSYSWRK subsystems that are
| already waiting for the named program.
| Ÿ Incoming IP port number corresponds to one of the
| service name-description-port-numbers documented in
| AS/400 Client Access Host Servers. This also applies to
| the pre-started jobs for the QSERVER, QCMN, and
| QSYSWRK subsystems that are already waiting for the
| assigned IP port number.
| Ÿ Incoming IPX socket number corresponds to one of the
| service name-description-port-number documented in
| AS/400 Client Access Host Servers. This also applies to
| the pre-started jobs for the QSERVER, QCMN, and
| QSYSWRK subsystems that are already waiting for the
| assigned IPX port number.

Chapter 7. Performance Reports—Manager Feature 7-139

| Ÿ Incoming 5250 display emulation jobs that come from
| APPC data streams sent by 5250 emulation under OS/2
| Communications Manager or WARP equivalent.
D Target distributed data management (DDM) server
| I Interactive. For the Component Report, this includes twinaxial
| data link control (TDLC), 5250 remote workstation, and 3270
| remote workstation. For the Transaction Report, this includes
| twinaxial data link control (TDLC), 5250 remote workstation,
| 3270 remote workstation, SNA pass-through, and 5250
| TELNET
L Licensed Internal Code Task
M Subsystem monitor
| P SNA pass-through and 5250 TELNET pass-through. On the
| Transaction Report, these jobs appear as I (interactive).
R Spool reader
S System
| W Spool writer, which includes the spool write job, and if
| Advanced Function Printing (AFP) is specified, the print
| driver job.
| WP Spool print driver (Transaction only)
X Start the system
The possible job subtypes are:
D Batch immediate job
E Evoke (communications batch)
J Pre-start job
P Print driver job
T Multiple requester terminal (MRT) (System/36 environment
only)
3 System/36
Notes:
| 1. Job subtypes do not appear on the Component Report.
2. If the job type is blank or you want to reassign it, use the Change
Job Type (CHGJOBTYP) command to assign an appropriate job
type.
Type (System, Transaction, Job Interval) One of the transaction types listed in
the description of the DTNTY field in Table 7-9 on page 7-96.
(System) The disk type.
(Transaction) The type and subtype of the job.
(Transaction) For the Seize/Lock Conflicts by Object section, the type
of seize/lock conflict.

7-140 Performance Tools V4R2

 Unit (System, Component, Resource Interval) The number assigned by the
system to identify a specific disk unit or arm. An ‘A’ or ‘B’ following the
unit number indicates that the disk unit is mirrored. (For example, 0001A
and 0001B are a mirrored pair.)
Unit Name The resource name of the disk arm.
User ID (System, Component, Transaction, Job Interval, Pool) The list of users
selected to be included (SLTUSRID parameter) or excluded
(OMTUSRID parameter).

User Name (Component, Transaction, Job Interval, Batch Job Trace) Name of the
user involved (submitted the job, had a conflict, and so on.)
| User Name/Thread (Transaction) If the job information contains a secondary
| thread, then this column shows the thread identifier. If the job informa-
| tion does not contain a secondary thread, then the column shows the
| user name.
| The system assigns the thread number to a job as follows:
| Ÿ The system assigns thread IDs sequentially. When a job is started
| that uses a job structure that was previously active, the thread ID
| that is assigned to the initial thread is the next number in the
| sequence.
| Ÿ The first thread of a job is assigned a number.
| Ÿ Any additional threads from the same job are assigned a number
| that is incremented by 1. For example:
| Job Name User Name/ Job Number
| Thread
| QJVACMDSRV SMITH ð23416
| QJVACMDSRV ððððððð6 ð23416
| QJVACMDSRV ððððððð7 ð23416
| QJVACMDSRV ððððððð8 ð23416
| A thread value greater than 1 does not necessarily mean the job
| has had that many threads active at the same time. To determine
| how many threads are currently active for the same job, use the
| WRKACTJOB, WRKSBSJOB, or WRKUSERJOB commands to find
| the multiple three-part identifiers with the same job name.
User Starts (Component) The number of start journal operations initiated by the
user.
User Stops (Component) The number of stop journal operations initiated by the
user.
User Total (Component) The total number of journal deposits resulting from
system-journaled objects.
Util (Component, Resource Interval) The percent of utilization for each local
work station, disk, or communications IOP, controller, or drive.
Note: The system-wide average utilization does not include data for
mirrored arms in measurement intervals for which such intervals
are either in resuming or suspended status.
Util 2 (Component, Resource) Utilization of co-processor.

Chapter 7. Performance Reports—Manager Feature 7-141

 Value (Transaction) For the Individual Transaction Statistics section of the Job
Summary report, it is the value of the data being compared for the trans-
action. For the Longest Seize/Lock Conflicts section, it is the number of
seconds in which the seize or lock conflict occurred.
Verify (Component) Number of verify exceptions per second. Verify exceptions
occur when a pointer needs to be resolved, when blocked MI
instructions are used at security levels 10, 20, or 30, and when an unre-
solved symbolic name is called.
This count could be very high, even under normal system operation.
Use the count as a monitor. If there are large variations or changes,
explore these variations in more detail.
W-I Wait/Tns (Transaction) The average time, in seconds, of wait-to-ineligible time
per transaction. This value is an indication of what effect the activity
level has on response time. If this value is low, the number of wait-to-
ineligible transitions probably has little effect on response time. If the
value is high, adding additional interactive pool storage and increasing
the interactive pool activity level should improve response time. If you
are unable to increase the interactive pool storage (due to limited avail-
able storage), increasing the activity level may also improve response
time. However, increasing the activity level might result in excessive
faulting within the storage pool.
Wait Code (Transaction) The job state transition that causes the trace record to be
produced. The values can be as follows:
EVT Event Wait. A long wait that occurs when waiting on a
message queue.
EOTn End of transaction for transaction for type n. 3

EORn End of response time for transaction n. 3

HDW Hold Wait (job suspended or system request).

LKRL Lock Released. The job released a lock it had on the object
named on the next detail line of the report (OBJECT --). The
job that was waiting for the object is named on this line
(WAITER --) along with the amount of time the job spent
waiting for the lock to be released.
LKW Lock Wait. If there are a number of these, or you see entries
with a significant length of time in the ACTIVE/RSP* column,
additional analysis is necessary. The LKWT report lines that
precede this LKW report line show you what object is being
waited on, and who has the object.
LKWT Lock Conflict Wait. The job is waiting on a lock conflict. The
time (*/ time /*) is the duration of the lock conflict and,
though not equal to the LKW time, should be very close to it.
The holder of the lock is named at the right of the report line
(HOLDER --). The object being locked is named on the next
report line (OBJECT --).

3 These codes are in the wait code column, but they are not wait codes. They indicate transaction boundary trace records. For more
information see Chapter 8, “Transaction Boundaries—Manager Feature” on page 8-1.

7-142 Performance Tools V4R2

 SOTn Start of transaction n. 3

SWX Short Wait Extended. The short wait has exceeded a

2-second limit and the system has put the transaction into a
long wait. This long wait must be charged to the transaction
response time. In other words, this active-to-wait transaction
does not reflect a transaction boundary.
SZRL Seize/Lock Released. The job released a lock it had on the
object named on the next detail line of the report
(OBJECT --). The job that was waiting for the object is
named on this line (WAITER --) along with the amount of
time the job spent waiting for the seize conflict to be
released.
SZWT Seize/Lock Conflict Wait. The job is waiting on a seize/lock
conflict. The time (*/ time /*) is the duration of the seize/lock
conflict, and is included in the active time that follows it on
the report. The holder of the lock is named at the right of the
report line (HOLDER --). The object being held is named on
the next report line (OBJECT --).
TSE Time Slice End. The program shown in the stack entry
labeled LAST is the program that went to time slice end.
Every time a job uses 0.5 seconds of CPU time (0.2 seconds
on the faster processors) between long waits, the system
checks if there are jobs of equal priority on the CPU queue. If
there are, then the next job with equal priority is granted the
CPU and the interrupted job is moved to the queue as the
last of equal priority. The job, however, retains its activity
level. This is an internal time slice end.
When a job reaches the external time slice value, there can
be a job state transition from active to ineligible if another job
is waiting for an activity level. When a job is forced out of its
activity level, its pages are liable to be stolen by other jobs,
and cause additional I/O when the job regains an activity
level.
The IBM-supplied default values of 2 seconds for interactive
jobs and 5 seconds for batch jobs may often be too high,
especially for the high-end processors. As an initial value, set
the time slice at 3 times the average CPU seconds per trans-
action.
WTO Wait Timed Out. The job has exceeded the wait time-out limit
defined for a wait (such as a wait on a lock, a message
queue, or a record).
Wait-Inel (System, Component) Average number of wait-to-ineligible job state
transitions per minute.
Work Station Controller (Resource Interval) The name of the remote work station
controller.
Working set size (KB) (System) An estimate of the main storage size, in kilobytes,
required by a single interactive job running the transactions described by
this report.

Chapter 7. Performance Reports—Manager Feature 7-143

 Writes per Second (Resource Interval) The average number of disk write oper-
ations performed per second by the disk arm.
WTO (Transaction) Listed in the Wait Code column, Wait Timed Out. The job
has exceeded the wait time-out limit defined for a wait (such as a wait
on a lock, a message queue, or a record).
0.0-1.0 (Component, Resource Interval) The number of times the response time
was between 0 and 1 second.
1.0-2.0 (Component, Resource Interval) The number of times the response time
was between 1 and 2 seconds.
2.0-4.0 (Component, Resource Interval) The number of times the response time
was between 2 and 4 seconds.
4.0-8.0 (Component, Resource Interval) The number of times the response time
was between 4 and 8 seconds.

7-144 Performance Tools V4R2

Chapter 8. Transaction Boundaries—Manager Feature
A transaction is a basic unit of work done on a system. The type of work varies,
depending on what kind of work it is or who is doing the work. Performance Tools
reports capture information about many kinds of system transactions; you can then
use these reports to analyze system performance.

When the Transaction Report counts transactions, it uses only state transactions.
For example, when a job goes from wait to active state, this marks the beginning of
a transaction. When a job goes from active to wait, the transaction is considered
ended. For the display I/O transactions and the data queue transactions, you can
specify values *DI and *DQ. These values use existing transaction boundary trace
records to count transactions instead of the wait-to-active state transition.

This chapter provides information about the following types of system transactions:
Ÿ Display I/O information
Ÿ SNA performance measurements
Ÿ APPN control point performance
Ÿ APPC protocol
Ÿ Performance measurement and SNADS
Ÿ SNADS sample data
Ÿ SNADS performance notes
Ÿ Pass-through
Ÿ Licensed Internal Code server
Ÿ Data queue transactions

Display I/O Transaction Boundary Information

The transaction boundary information in Figure 8-1 shows how a display I/O trans-
action uses system resources by showing the relationship between transaction
response time and resource usage time.

Workstation AS/400

6 5 4

1 2 3
Application

RV2S090-1

Figure 8-1. Example: Display I/O Transaction

The numbers 1 through 6 in the following list refer to the same numbers in
Figure 8-1.
1 The user presses the Enter key or a function key. This begins the
response time period perceived by the user. However, the system does
not recognize the beginning of the transaction until step 2.

 Copyright IBM Corp. 1998 8-1

 Delays are typical on a remote communication line. They are dependent
on the following:
Ÿ The amount of current data traffic to and from other work stations on
the line.
Ÿ How frequently the system polls the control unit for input data.

2 Start of Transaction (SOT) 1

Identifies the beginning of the System Measured Response Time. Work-

station I/O Management (WSIOM) processes input from the display
station. This also represents the beginning of application-input queuing
time.
This is a trace data point.
3 Start of Resource Utilization Time (SOR)
The application must issue an input operation or accept an input opera-
tion. An application program receives the data from WSIOM and begins
using system resources to process the transaction. The application-input
queuing time ends at this point. Normally, application-input queuing time,
like activity-level waiting time, is only a few milliseconds.
4 End of Resource Utilization Time (EOR) 1

The application program completes using system resources. This

normally coincides with the End of Transaction (EOT).
At this point, the program has performed an operation that causes work
station I/O to send data to the display station. The following user
program operations cause the data to be sent to the display station:
Ÿ Read or invite input operation following one or more output oper-
ations with the defer write (DFRWRT) parameter set to *YES in the
display file description.
Ÿ Output operation with DFRWRT(*NO) in the display file description.
Ÿ Output operation with the DDS INVITE keyword.
Ÿ Combined output/input operation. For example, an EXFMT operation
in an RPG/400 program and a SNDRCVF command in a control lan-
guage (CL) program.
Ÿ End of program.
This is a sample data point.
5 End of Transaction (EOT) 1

The end of the System Measured Response Time. The next transaction
may begin. Resource usage by the transaction is measured at this point.
This may coincide with the End of Resource Utilization Time (EOR). Any
Active-Wait transition is included here.
This is a trace data point.
6 System response displayed to user.

1 The SOT, EOR, and EOT abbreviations appear on the Transition Report. For an example of a Transition Report, see Figure 7-49
on page 7-62.

8-2 Performance Tools V4R2

 1→6 Display I/O Transaction Path
The complete path taken by the transaction. This is the time from when
the user presses the Enter key or a function key to the time when the
user receives a response. This is the user's perception of the response
time.
2→3 Application Input Queuing Time
This is the time the input data waits before the system resources are
made available to it. Examples are input data waiting on:
Ÿ an activity level
Ÿ the program to issue an input operation
Ÿ the program to accept input.
The total application queuing time (in hundredths of a second) is stored
in the JBAIQT field in the QAPMJOBS file. The number of application
queuing transactions is stored in the JBNAIQ field in the QAPMJOBS
file. These fields are also updated by data queue transactions.
3→4 Transaction Resource Usage
The period when system resources are used for processing including
periods of waiting, such as object seize/lock conflicts and resource
queuing.
The total resource usage time (in seconds) is stored in the JBRUT field
in the QAPMJOBS file. The number of resource usage transactions is
stored in the JBNRU field in the QAPMJOBS file. These fields are also
updated by data queue transactions.
2→5 System Response Time
The total transaction time (in seconds) is stored in the JBRSP field in
the QAPMJOBS file. The number of transactions (5250 only) is stored in
the JBNTR field in the QAPMJOBS file. These fields are also updated
by Client Access shared folders transactions and by pass-through trans-
actions.
1→2 Components of response time are not recorded by the system.
5→6 Components of response time are not recorded by the system.

SNA Performance Measurements

The SNA performance measurements provide a different set of internal perform-
ance data for each APPC and host controller description. These measurements
include the activity created by attached device descriptions and APPN intermediate
sessions.

Performance data is collected for a controller description only after the controller is
varied on and at least one connection has been established with the adjacent
system. Performance data is not collected after the controller description is varied
off.

The QAPMSNA file contains the SNA performance measurements. The fields in the
QAPMSNA file are categorized as follows:
Ÿ Correlation fields

Chapter 8. Transaction Boundaries—Manager Feature 8-3

 Ÿ Connection fields
Ÿ Device description fields
Ÿ T2 station I/O manager task fields
Ÿ Session traffic fields

Correlation Fields
Correlation fields include external configuration names and internal task names that
allow the performance measurements to be correlated to other parts of the system.

Correlating the SNA performance measurements with other parts of the system is
important. The following correlation fields are defined:
SCTLNM Names the APPC or host controller description.
SLINNM Names the line description that is attached to the controller description.
If *LOCAL is specified for the link type parameter on an APPC controller
description, this field is blank.
STSKNM Identifies the T2 station IOM task that provides services for the controller
description. The QAPMJOBS performance monitor file contains informa-
tion about processing unit use and disk unit accesses for this task.
SLIOMT Identifies the line IOM task that provides services for the line description.
The QAPMJOBS performance monitor file contains information about
processing unit use and disk unit accesses for this task.
Note: Because the line IOM task could service multiple station IOM
tasks, the processing unit use and disk unit access data may not
be attributed to a single station IOM task or controller
description. For example, multiple controller descriptions are
often attached to a single LAN line description.
SACPNM Names the adjacent control point. If the controller description is not
APPN capable, this field may be blank. The adjacent CP name can be
used to correlate with data displayed by the Display APPN Information
(DSPAPPNINF) command.
SANWID Names the adjacent network ID. The adjacent network ID can be used
to correlate with data displayed by the DSPAPPNINF command.
SAPPN Indicates whether or not the controller description is APPN capable. If
the system uses the APPN support, additional performance measure-
ments can be found in the QAPMAPPN performance monitor file.
SCTYP Indicates whether the controller description is an APPC or host con-
troller.

Connection Fields
Connection fields measure the frequency with which connections are established
with the adjacent system.

A connection is established with the adjacent system when the status of the con-
troller description goes from varied off or vary on pending to varied on or active.
You can view this status using the Work with Configuration Status (WRKCFGSTS)
command.

8-4 Performance Tools V4R2

 On a non-switched line, the connection is established after the line and controller
description are varied on, assuming the adjacent system is ready to establish the
connection. The non-switched connection remains until the controller is varied off,
an irrecoverable line error occurs, or the adjacent system drops the connection.

On a switched line, the connection is not established until a communications

program needs to use the connection (for example, the program acquires a
session). The switched connection is usually dropped after the connection has been
inactive (for example, all sessions are unbound) for a period of time.

The following connection fields are defined:

STLLBU Indicates the date and time when the most recent connection was estab-
lished.
SNLBU Indicates the number of connections that were established with the adja-
cent system in the time interval. Dropping and re-establishing a con-
nection frequently can degrade performance. Using switched lines, you
may frequently re-connect if you have an inappropriate Switched discon-
nect parameter value on the controller description or irrecoverable line
errors.

Device Description Fields

Device description fields measure device-related activity. APPN automatically
creates, varies on, and deletes devices.

The following device description fields are defined:

STACVO Indicates the cumulative elapsed time in milliseconds required to auto-
matically create and vary on APPN devices.
SNACVO Indicates the number of APPN devices that were automatically created
or varied on.
SNADD Indicates the number of APPN devices that were automatically deleted.
Note: If devices are automatically created or deleted excessively, your system's
performance can be degraded. Increasing the number of minutes specified
on the Autodelete device parameter on the controller description reduces
the frequency with which APPN automatically deletes devices.

T2 Station I/O Manager Task Fields

These task fields give an estimate of how much work is being done by the T2 (PU
type 2) station I/O manager task.

The T2 SIOM task provides services for the controller description. The processing
unit utilization and disk unit accesses for the T2 SIOM task are contained in the
QAPMJOBS performance monitor file. A description of that file can be found in the
Work Management book.

The following T2 SIOM task fields are defined:

SNWAIN Indicates the number of internal task messages that are received by the
T2 SIOM task. This field is an approximation of the amount of work per-
formed by the T2 SIOM task.
SNWAOU Indicates the number of internal task messages that are sent by the T2
SIOM task.

Chapter 8. Transaction Boundaries—Manager Feature 8-5

Session Traffic Fields
Session traffic fields measure the sending and receiving of session traffic. A sepa-
rate set of identical session traffic fields is collected for each session type and pri-
ority level combination.

There are two session types: end point and intermediate sessions. End point
session traffic is created by the following device types:
Ÿ APPC devices
Ÿ Host devices (for example: 3270 emulation, RJE)
Ÿ DHCF display devices
Ÿ NRF display and printer devices

Intermediate session traffic is created by the following:

Ÿ APPN intermediate sessions
Ÿ SNA pass-through devices

There are four priority levels: network, high, medium and low. Network priority
session traffic is created by the following:
Ÿ APPN
Ÿ SNA change to the number of sessions
Ÿ Alert support

High priority session traffic is created by the following:

Ÿ APPC devices
Ÿ APPN intermediate sessions

Medium priority session traffic is created by the following:

Ÿ APPC devices
Ÿ Host devices (for example, 3270 emulation, RJE)
Ÿ DHCF display devices
Ÿ NRF display and printer devices
Ÿ SNA pass-through devices
Ÿ APPN intermediate sessions

Low priority session traffic is created by the following:

Ÿ APPC devices
Ÿ APPN intermediate sessions

Therefore eight different sets of session traffic fields are collected.

The first two characters of the session traffic field name represent the session type
and priority level. The first character specifies the session type:
E End point
I Intermediate

8-6 Performance Tools V4R2

 The second character specifies the priority level:
N Network
H High
M Medium
L Low

The remaining four characters represent the function of the field. Figure 8-2 on
page 8-8 shows the layout of the session traffic fields in the QAPMSNA file.
Note: Throughout the remainder of this section, the first two characters of the
session traffic fields are replaced with the prefix tp to generically refer to
any session type and priority level combination.

Number of Sessions Started and Ended

The tpNSS and tpNSE fields count the number of sessions that are started and
ended, respectively. A session starts when the positive response to an SNA bind
command is sent or received. A session ends when an SNA unbind command is
sent or received, or the session is abnormally ended (for example, the line fails).
Starting and ending sessions can cause significant system overhead.

Number of Brackets Started and Ended

The tpNBB and tpNEB fields count the number of SNA brackets that are started
and ended. For sessions that are not LU 6.2, the delimiters for the start and end of
a bracket are the bind and unbind command. For LU 6.2, the delimiters for the start
and end of a bracket are the begin bracket indicator (BBI) and conditional end
bracket indicator (CEBI) in the request header (RH). An LU 6.2 bracket is roughly
equivalent to a conversation that is started when a program issues an ICF evoke
operation or Common Programming Interface Communications (CPI-C) allocate
verb and ends when the Common Programming Interface Communications program
issues an ICF detach operation or Common Programming Interface Communi-
cations (CPI-C) deallocate verb. An example is a DSPT (display station pass-
through) or SNADS session.

Sending Data
The SNA processing required to send data can be classified into the following
stages:
Ÿ Session-level pacing
Ÿ Internal session-level pacing
Ÿ Transmission priority
Ÿ Line transmission

Session-Level Pacing
Session-level pacing is a technique that allows a receiving session to control the
rate at which it receives request units on the normal flow. It is used primarily to
prevent a receiver with unprocessed requests from overloading because the sender
can create requests faster than the receiver can process them.

The following session-level pacing fields are defined:

Chapter 8. Transaction Boundaries—Manager Feature 8-7

 ENffff
.
. End Point Session/Network Priority
.
EHffff
.
. End Point Session/High Priority
.
EMffff
.
. End Point Session/Medium Priority
.

ELffff
.
. End Point Session/Low Priority
.

INffff
.
. Intermediate Session/Network Priority
.
IHffff
.
. Intermediate Session/High Priority
.
IMffff
.
. Intermediate Session/Medium Priority
.
ILffff
.
. Intermediate Session/Low Priority
.

ffff = 1 to 4 character functional name RV2S070-0

Figure 8-2. Layout of Session Traffic Fields

tpSPWT Specifies the cumulative amount of time that application data was
waiting for a pacing response to be received.
tpSPNW Specifies the number of times that application data was waiting for a
pacing response to be received.
tpSPPW Specifies the total number of pacing windows, which is the potential
number of times that application data could have waited for a pacing
response to be received.
tpSPWS Specifies the cumulative pacing window size.

The following information can be derived from the session-level pacing fields:
Ÿ The average amount of time spent waiting for a pacing response to be received
is: tpSPWT/tpSPNW.
Ÿ The percentage of times application data waited for a pacing response to arrive
is: (tpSPNW*100)/tpSPPW.
Ÿ The average pacing window size is: tpSPWS/tpSPPW.

If excessive waiting is caused by session-level pacing, the OUTPACING (local

system) and INPACING (remote system) parameters in the mode description may
need to be increased. However, if the average pacing window size is 7 or more, the
excessive waiting may be caused by a slow remote system or a slow remote
program.

8-8 Performance Tools V4R2

Internal Session-Level Pacing
For APPN and APPC sessions that are adaptively paced, internal session-level
pacing is used to limit the amount of bandwidth used by a particular session. It only
controls internal flow and does not have any external line flows. A sending session
is allowed to transmit a limited number of request units and is not allowed to
transmit additional request units until a request unit is successfully delivered to the
adjacent system.

The INPACING and OUTPACING parameters in the mode description are used to
calculate the limit. The limit used for a given session is (2*n)-1, where n is the
INPACING or OUTPACING parameter. On a slow speed line, it may be necessary
to configure a small limit for batch traffic and a larger limit for interactive traffic to
ensure acceptable interactive response time.

The following internal session-level pacing fields are defined:

tpIPWT Specifies the cumulative amount of time that application data was
waiting, because of internal session-level pacing.
tpIPNW Specifies the number of times that application data was waiting,
because of internal session-level pacing.

The following information can be derived from the internal session-level pacing
fields:
Ÿ The average amount of time spent waiting because of internal session-level
pacing is: tpIPWT/tpIPNW.

If excessive waiting is caused by internal session-level pacing and it is not desir-

able to limit the amount of bandwidth used, the OUTPACING and INPACING
parameters in the mode description may need to be increased.

Transmission Priority
Transmission priority determines the criteria for being selected for transmission to
the adjacent system by allowing different priority levels to be assigned to session
traffic. Three user-defined priorities are defined: high, medium and low. Network pri-
ority is reserved for APPN and SNA control traffic. Interactive traffic is typically
assigned high priority, and batch traffic is typically assigned medium or low priority.

The following transmission priority fields are defined:

tpQNRE Specifies the number of request units that entered the transmission pri-
ority queue.
tpQLRE Specifies the cumulative length of data that entered the transmission pri-
ority queue.
tpQNRL Specifies the number of request units leaving the transmission priority
queue to be sent to data link control for transmission to the adjacent
system.
tpQLRL Specifies the cumulative length of data leaving the transmission priority
queue to be sent to data link control for transmission to the adjacent
system.
tpQTRR Specifies the cumulative amount of time that request units waited in the
transmission priority queue.

Chapter 8. Transaction Boundaries—Manager Feature 8-9

 The following information can be derived from the transmission priority fields:
Ÿ The average length of a request unit entering the transmission priority queue is:
tpQLRE/tpQNRE.
Ÿ The average length of a request unit leaving the transmission priority queue is:
tpQLRL/tpQNRL.
Ÿ The average amount of time a request unit waited in a transmission priority
queue is: tpQTRR/tpQNRL.

Excessive waiting in a transmission priority queue may occur if:

Ÿ The data is preempted by higher priority data
Ÿ The line is a slow speed line
Ÿ Frequent retransmissions are required due to an error-prone line
Note: The average wait time for higher priority data should typically be less than
lower priority data.

Line Transmission
Performance data is collected to quantify the amount of time required to success-
fully transmit data to the adjacent system. This measurement period begins after
the data leaves the transmission priority queue and ends when the data is success-
fully delivered to the adjacent system.

The following line transmission fields are defined:

tpNRUD Specifies the number of request units delivered to the adjacent system.
tpLRUD Specifies the cumulative length of data delivered to the adjacent system.
tpTRUD Specifies the cumulative amount of time in milliseconds to deliver data
to the adjacent system.

The following information can be derived from the line transmission fields:
Ÿ The average length of a delivered request unit is: tpLRUD/tpNRUD.
Ÿ The average amount of time to deliver a request unit is: tpTRUD/tpNRUD.
Note: This data does not provide an accurate measurement of line utilization
because only a portion of the data being transmitted on the line is meas-
ured.

Receiving Data
Performance data is collected to record the number of request units and cumulative
length of data that is received.

The following fields are defined:

tpNRUR Specifies the number of request units received from the adjacent
system.
tpLRUR Specifies the cumulative length of request units received from the adja-
cent system.

The following information can be derived:

Ÿ The average length of a received request unit is: tpLRUR/tpNRUR.

8-10 Performance Tools V4R2

Intermediate Session Traffic Work Load
The work load caused by intermediate session traffic can be estimated because a
separate set of session traffic fields is defined for intermediate and end point
session traffic. The values from the corresponding intermediate session field for
each priority level can be added together to determine the overall system work
load. For example, INNRUR+IHNRUR+IMNRUR+ILNRUR is the total number of
request units received on all intermediate sessions.

The intermediate session work load can be reduced by decreasing the maximum
intermediate sessions parameter in the network attributes, or decreasing the pacing
counts configured on the end point systems. The AS/400 system configures the
pacing counts on the OUTPACING and INPACING parameters on the mode
description.

Control Traffic Work Load

The work load caused by control traffic can be estimated by examining the network
priority session traffic fields. Network priority is reserved for control traffic. High,
medium and low priority are used by user traffic. APPN control traffic uses end
point sessions only. Operations that change the number of sessions and alert
control traffic may use both end point and intermediate sessions.

Comparing Different Priority Levels

The SNA performance measurements allow the proportion of network, high,
medium and low priority traffic to be analyzed. High priority is typically reserved for
interactive jobs that require good response time and medium and low priority is
assigned to batch jobs.

The priority level is configured in the Transmission priority parameter on the class-
of-service description.

APPN Control Point Performance Measurements

The primary purpose of the APPN control point is to allow applications on one
system to dynamically establish sessions with applications on another system.
Because of the dynamics involved in APPN, there are many work activities that an
APPN network node or end node needs to perform to maintain the information
required to establish sessions. Refer to the APPN Support book for background
information and terminology for APPN.

The APPN performance measurements provide a granular breakdown of these

work activities. The file QAPMAPPN does not contain any data regarding proc-
essing unit utilization or disk unit accesses. Information regarding processing unit
utilization and disk unit accesses for the tasks that perform APPN functions can be
found in the QAPMJOBS performance monitor file. A performance analyst can then
determine the activities that APPN was performing (to better understand the
resource utilization found in the QAPMJOBS file). Use the QAPMAPPN file with
the QAPMJOBS file to determine the effect of APPN functions on a system's per-
formance.

The QAPMAPPN file does not contain any information regarding session traffic.
APPN session traffic data is maintained in the file QAPMSNA. There is a set of
measurements for each active controller description on the system. The

Chapter 8. Transaction Boundaries—Manager Feature 8-11

 QAPMSNA file data provides performance measurements to give a breakdown of
intermediate routing and session endpoint traffic. The session traffic data provides a
breakdown based on the different transmission priorities that can be used. Refer to
“Session Traffic Fields” on page 8-6 for more information on session traffic.

APPN Work Activities

The following are the various APPN work activities:
Ÿ Topology maintenance
Ÿ Directory services registration and deletion requests
Ÿ Configuration changes
Ÿ Control point session activation and deactivation
Ÿ Control point presentation services
Ÿ Session setup activities

Topology Maintenance
These work activities maintain the APPN topology database. The APPN topology
database allows routes through the APPN network to be calculated, based on a
class-of-service selected by the user initiating a session. You can display how often
the topology database is viewed using the Display APPN information
(DSPAPPNINF) command.

Topology maintenance can account for a large amount of processing unit and disk
unit accesses. In general, the amount of resource required for topology mainte-
nance grows as networks become larger and as the APPN network becomes
unstable. The instability of a network is caused by frequently activating and deacti-
vating transmission groups or by having line failures and systems failures in the
network. Following are some key terms regarding topology maintenance and a
short explanation of how these work activities affect performance:
Transmission group (TG) update A TG update occurs when a controller
description on the local system has a status change (for example, it
changes from inactive to active). When the TG defines a connection
between two network nodes, this causes the local system to send a
topology database update.
Topology database update (TDU) A TDU is the device used to broadcast a status
change of a resource in the intermediate routing portion of an APPN
network. An AS/400 system sends TDUs for several different reasons. A
TDU is a general data stream (GDS) variable that can block information
about multiple resources into a single entity. Thus, if the topology
routing services (TRS) component receives multiple TG updates, it can
block these together into a single TDU. TDUs are distributed to every
network node in the APPN network that is connected to the remaining
systems in the network using control point sessions.
Node congestion updates These occur when a network node has a status change
in its capacity for performing intermediate routing. On the AS/400
system, node congestion is based simply on the number of intermediate
sessions that are currently active. Node congestion updates cause the
local system to send TDUs.

8-12 Performance Tools V4R2

Received TDUs APPN network nodes receive TDUs as well as send them. If TDUs
are received frequently, this can degrade system performance. The
counts maintained for received TDUs provide a breakdown between new
and old resources.
One count tracks the number of times the most frequently used node
was included in received TDUs during an interval. When many TDUs
are received and the same node is always listed in the APPN perform-
ance data, this could indicate a configuration problem where the listed
node has updates sent continuously. This situation can have a serious
effect on performance. For more details see the APPN Support book.
Initial topology exchange An initial topology exchange is an examination of the
resources in the intermediate routing portion of the topology database
that occurs any time control point sessions are established between two
network nodes. Any of the resources that have changed status or for
which a TDU has been received are sent to the partner network node in
a TDU. If one of the systems has refreshed its topology database or if
this is the first initial topology exchange with the partner network node
since performing a system IPL, the entire intermediate routing portion of
the APPN topology database is sent.
Obsolete Topology Entry Removal Once every 24 hours the topology database is
examined to determine if any entries in the local system have not been
updated in the past 15 days. The entries that have not been updated
are deleted. If the local system is a network node, the system sends a
TDU once every five days (so other nodes do not delete the local
system from their topology database).
Display APPN information (DSPAPPNINF) Each time the Display APPN Informa-
tion (DSPAPPNINF) command is run with *TOPOLOGY specified for the
information type (INFTYPE) parameter, every entry in the APPN
topology database is examined. In networks with a large amount of
topology, this can account for a significant number of read operations
from disk units.
Note: This does not directly affect topology maintenance.

Directory Services Registration and Deletion Requests

APPN end nodes register and delete their local location names with their network
node server. For end nodes, these configuration change requests are tracked
because configuration changes cause an end node to send the registration and
deletion requests.

These measurements are maintained on a network node to show the effort involved
in processing received location registration and deletion requests from attached end
nodes. Various conditions cause an end node to send in registration and deletion
requests (for example, the activation of a control point session, or a configuration
change). The conditions that cause an end node to send these requests can be
found in the configuration and control point session performance measurements.
Multiple locations can be included in a single registration request (such as when an
end node is registering all of its locations following control point session activation).

In general, registration and deletion requests should not cause a performance

burden on a network node because this type of information is not distributed to
every network node in the network.

Chapter 8. Transaction Boundaries—Manager Feature 8-13

 Configuration Changes
These work activities cause the APPN control point to update configuration informa-
tion and can at times cause an update to be sent to one or more systems. The
activities included in this section are:
Ÿ Change network attributes processing
Ÿ Local location list updates
Ÿ Remote location list updates
Ÿ Mode description updates
Ÿ Class-of-service description
Change network attributes (CHGNETA) All of the APPN tasks are involved in
processing a Change Network Attributes (CHGNETA) command so local
information can be updated. The topology routing services task sends
out a TDU if the local system is a network node and the route addition
resistance (RAR) is changed. Changing the local system node type,
local network ID, or local control point name can cause the APPN
topology and APPN directory databases to be deleted. This can indi-
rectly affect performance (because additional processing is required to
reconstruct these databases).
APPN local location list updates These cause a local location to be added to or
deleted from the APPN directory database. If the local system is an end
node with a CP-CP session established to a network node server, these
cause a registration or deletion request to be sent to the server.
APPN remote location list updates These cause a remote location to be added to
or deleted from the APPN directory database.
Mode updates These cause the control point manager (CPMGR) task to update its
mode tables to reflect the addition, deletion, or update of a mode
description.
Class-of-service updates These cause the control point manager (CPMGR) and
topology routing services (TRS) tasks to update their class-of-service
tables to reflect the addition, deletion, or update of a class-of-service
description.

Control Point Session Activation and Deactivation

These work activities track the number of control point sessions that are started
and ended. These counts are classified as the locally controlled (contention winner)
CP-CP sessions or the remotely controlled (contention loser) CP-CP sessions.

There are various details associated with the activation and deactivation of control
point sessions. Contention winner control point sessions are primarily used for
sending data (TDUs, directory searches). Contention loser control point sessions
are used to receive control point data from other systems. The activation of a con-
tention winner CP session has many similarities to the activation of a user session.
The steps involved with the activation of a session (single hop route requests, acti-
vate route requests, device selection) are discussed in “Session Setup Work
Activity Details” on page 8-19.

When the local system is an end node and it activates a control point session to its
network node server, all of the local locations are registered with the network node.

8-14 Performance Tools V4R2

The control point session performance measurements also provide counts of the
number of control point sessions that are currently active. Variations in these
counts can help explain changes in resource utilization over different time intervals.
Because the number of active control point sessions can affect the number of
systems involved in TDU search procedures, the APPN tasks that provide the
majority of the function for control point session activation and deactivation are
control point manager (CPMGR) and control point presentation services (CPPS).

Control Point Presentation Services (CPPS)

Control point presentation services handles all of the data transfer that occurs on
the CP-CP sessions for the other APPN tasks. Analysis of these counts provides a
summary of the types of activities in which the APPN control point is involved over
a given set of time intervals. This information is classified into data sent and data
received for the various APPN transaction programs. These transaction programs
are:
Ÿ Control point (CP) capabilities
Ÿ Topology database update
Ÿ Directory services for search processing
Ÿ Registration/deletion

To best utilize the APPN performance measurements, the control point presentation
services measurements (CPPS) should be analyzed first. These measurements
provide a summary of the data traffic over the control point sessions for the various
APPN transaction programs. This summary discusses on the correct APPN work
activities and isolates any APPN performance problems. For example, if a time
interval shows a high number of directory services transactions and only a limited
number of topology database updates, the session setup measurements, instead of
the topology maintenance measurements, should be checked.

The control point presentation services measurements provide:

Ÿ The number of data received requests (data received from other network
locations that are directly connected to each other).
Ÿ The amount of data received.
Ÿ The number of send data requests (data sent to other network locations that
are directly connected to each other).
Ÿ The amount of data sent.

This information is provided for all of the different APPN transaction programs.
These transaction programs are:
Control point (CP) capabilities Used to send and receive control point capabilities
to adjacent systems immediately after activating control point sessions.
In general, running CP capability transaction programs should only
slightly affect system performance.
Topology database update Used to send and receive TDUs. TDUs are sent on
contention winner CP sessions and received on contention loser CP
sessions. TDUs can significantly affect performance for network nodes.
If the CPPS measurements seem abnormally high (compared to other
time intervals), check the topology maintenance data to determine the
cause of the increase.

Chapter 8. Transaction Boundaries—Manager Feature 8-15

 Directory services (DS) Used to send and receive search requests to other nodes
in the network. Search processing can significantly affect performance
for network nodes, but generally it only slightly affects end nodes.
Search requests and responses are sent asynchronously by the APPN
directory services task. If the CPPS measurements seem abnormally
high (compared to other time intervals), check the session setup per-
formance measurements to determine the cause of the increase.
Registration and deletion Used to send location registration and deletion requests
(from an end node to a network node server). In general, registration
and deletion requests should not significantly affect performance.
However, if a several requests are causing the CPPS and DS tasks to
increase processing unit use, check the directory services registration
and deletion requests measurements.

Session Setup Activities

These measurements provide information on the various steps that are taken by
the APPN control point to process session initiation requests. Because establishing
an APPN session is a highly distributed function, the measurements provide a
breakdown of the work activity details based on the role a system plays in estab-
lishing a session. For example, a network node performs certain functions to allow
local users to establish sessions. However, a network node also performs func-
tions to allow attached end nodes to establish sessions. The performance measure-
ments in each case are separated so you can find where the resources are used.

The activities performed and the resources used vary between APPN end nodes
and network nodes. Because APPN session setup is a function distributed between
multiple systems, it is necessary to classify session setup work into different work
activities.

Each of these work activities causes different types of work to run on the local
system. Refer to “Session Setup Work Activities” on page 8-17 for a description of
the session setup work activities. Refer to “Session Setup Work Activity Details” on
page 8-19 for a description of measurable work details associated with session
setup. Many of these details are common between the different work activities. The
sample data for session setup keeps separate counts and cumulative elapsed times
for the different work activities being performed.

Following is some terminology used to describe various work activities performed

by APPN. The definitions of these terms are based on the context of this dis-
cussion.
Bind command The request unit sent by a node to set up the LU-LU session.
Broadcast search A search sent by a network node and eventually forwarded to
every network node in the network (connected by one or more CP-CP
sessions).
Directed search A directory search sent by one network node to a single network
node. A network node sends this search if it has information in its direc-
tory database that indicates the destination network node was the pre-
vious owner or network node server of the end node that owns the
location. A directed search may flow through multiple network nodes
along the way (these are intermediate nodes on a directed search). Only
the destination of the directed search performs additional search logic.

8-16 Performance Tools V4R2

 Domain broadcast A search sent by a network node to its adjacent end nodes that
specified they want these searches. Currently, the only time an AS/400
system sends these to other AS/400 systems is in a multiple network
environment.
NNS(OLU) A network node serving an end node that is initiating a session setup
request.
NNS(DLU) A network node serving an end node that is the destination of a session
setup request.
One hop search A search sent from an end node to a network node or from a
network node to an end node. This search is only sent to a single node.

Session Setup Work Activities

Following is a list of the session setup work activities. Each work activity has
various detailed measurements associated with it. For each work activity, there is a
cumulative elapsed time measurement and a count of the number of times a given
work activity was successful. The start time and end time measured (accounting for
the cumulative elapsed time measurements) and the criteria for counting a given
work activity successful are listed with the work activity.
1. Locally started sessions (source)
Description: Sessions that are started on the local system, including explicit
session initiation requests by a user as well as internal session initiation
requests (for example, sessions started for session limit negotiation or alert
traffic).
Start: The system determines the APPC device description to use for session
initiation request.
End: The system provides information regarding the APPC device description
request. The information is either a list of devices or an error code.
Success: One or more device descriptions are returned to the operating
system.
2. Receiver of search requests (local system = EN)
Description: The local system, an end node, receives a search request from
its network node server.
Start: The directory services task receives a locate request from the CPPS
task.
End: Directory services (DS) returns the locate search response to the CPPS
task or the CP session when the system that sent the search fails.
Success: Directory services returns a positive response to the search request
it had received.
3. Search processing on for attached EN (local system = NN)
Description: The local system, a network node, has received a search request
from a served end node that is initiating a session. The local system is respon-
sible for searching for the target system and then calculating a route to the
destination control point.
Start: The DS task receives a request from CPPS and determines it is a
search request from a served end node.

Chapter 8. Transaction Boundaries—Manager Feature 8-17

 End: DS returns the locate search response to CPPS or the CP session with
the end node that sent the message indicating that the search had failed.
Success: DS sends a positive response to the locate search request, and
routing information is supplied to the end node.
4. Intermediate node on a directed search
Description: The local system, a network node, has received a directed search
request from another network node. The only functions that need to be per-
formed in this case are forwarding the search to the next hop of the route, and
also forwarding the search response to the system that had sent the search to
the local system.
Start: DS receives a directed search request from CPPS and realizes the local
system is not the destination network node.
End: DS returns the locate search response to the system it received the
search from or the CP session ends between the local system and either the
system that sent the search to the local system or the system that the local
system forwarded the search to.
Success: DS successfully sends the directed search, receives a positive
response, and successfully returns the directed search response to the system
that had originally sent the search request.
5. Destination NN on a directed search - NNS(DLU)
Description: The local system, a network node, has received a directed search
request from another network node. In this case, the local system is the target
of the directed search because the location being searched for had at one time
resided on the local system or on an end node that was being served by the
local system.
Start: DS receives a directed search request from CPPS and realizes the local
system is the destination network node.
End: DS returns a positive response to the system that had sent the search to
the local system or the CP session ends with that system.
Success: The response given by DS to the system that had sent the search to
the local system is positive.
6. Broadcast search received
Description: Broadcast searches are processed only by network nodes. When
the local system receives a broadcast search, it sends the search to all of its
adjacent network node partners and it also determines if the location being
searched for resides on the local system or on a served end node. Broadcast
searches are the most costly search types from a performance standpoint
because of the number of systems that are involved.
Start: DS receives a broadcast search request from CPPS.
End: DS has forwarded the search response to the system that sent the broad-
cast search to the local system and it has processed all search responses from
systems to which the local system forwarded the broadcast search.
Success: The response sent back to the system that sent the search to the
local system is positive.
7. NN processing a received search from a node in a non-AS/400 network

8-18 Performance Tools V4R2

 Description: This work activity tracks the number of searches processed that
are started by systems in a different APPN network (based on network identi-
fier). Only searches received from network nodes in another network are
counted. Note that only the systems on the boundaries of the network will
maintain these measurements.
Start: DS receives a search request from CPPS and determines it is from a
node in another network.
End: DS returns the locate search response to CPPS or the CP session with
the non-AS/400 node that sent the message indicating that the search had
failed.
Success: DS sends a positive response to the non-AS/400 node that sent the
search request to the local system.
8. Network node processing a received bind from a node in the AS/400 (local)
network without routing information
Description: The local system, a network node, is responsible for determining
the control point that the target system resides on, calculating a route to the
destination control point, and forwarding the bind on to the next hop of the
route.
Start: The control point receives a bind request from the session connector
manager (the part responsible for establishing intermediate sessions) and
determines that the request was received from a node in the AS/400 network
and that routing information does not exist.
End: The control point returns a response to the session connector manager
task.
Success: The request for the intermediate session initiation is returned with a
positive response (which means the link for the next hop of the route has been
located and is active).
Note: Work activities 9, 10, and 11 have the same start, end, and success
definitions as work activity 8. Work activities 10 and 11 do not require
any search processing or route computation processing.
9. NN processing a received bind from a node in a non-AS/400 network without
routing information
10. NN processing a received bind from a node in the AS/400 network with
routing information
11. NN processing a received bind from a node in a non-AS/400 network with
routing information

Session Setup Work Activity Details

For APPN, session setup involves details such as:
Ÿ Initial screening to determine if existing sessions may be used
Ÿ Directory search processing to determine which system in the network owns the
destination location desired
Ÿ Route selection to determine the optimal route based on class of service
through the network
Ÿ Switched link activation
Ÿ Device selection and/or creation of a new device

Chapter 8. Transaction Boundaries—Manager Feature 8-19

 These detailed measurements are stored separately for each distinct work activity
in which the measurement can occur. For example, the system can issue a directed
search as a result of several different work activities (in this case work activities 1,
3, 7, 8, and 9). Therefore, five different sets of detailed measurements for directed
searches are there so that the person analyzing the data can determine which
activities are causing the directed searches to run.

Following are explanations of some of the key terms of the work activity details:
Initial screening These measurements are functions performed by the location
manager and the control point manager tasks. These measurements
indicate how many new sessions need to be started (which require full
control point services to complete the request) and session requests sat-
isfied by using existing bound sessions. There are also measurements
to count session initiation requests that get pended by the control point.
The pending of session initiation requests improves performance
because the directory search, route selection, and switched link acti-
vation phases need only be done once for multiple session initiation
requests received.
Directory search processing This step involves determining the control point that
owns the target system of a session initiation request. The APPN parts
that are most affected by directory search processing are DS and
CPPS. The effect of search processing on performance is greater on a
network node than on an end node because of the various roles that a
network node can have in search processing. Because a network node
processes the first positive response it receives on a search request and
sends this to the search originator, a network node can still process a
search request after the work activity that started the request has com-
pleted. Even though the bind for a session may have already been sent,
a network node may still be processing subsequent search responses
received from other nodes. The directory search processing phase can
be an asynchronous process if searches are sent to other systems,
which can account for increased values in various cumulative elapsed
time measurements.
Route selection Route selection is carried out by the TRS task. There are different
types of routes that TRS calculates. A single hop route is done by an
end node (when an end node has not received routing information from
a network node server). A single hop route is also done for establishing
a control point session. Request route processing is done by network
nodes for establishing end-to-end routes based on a particular class-of-
service.
Switched link activation This processing is primarily carried out by the machine
services control point (MSCP) task. MSCP receives activate route
requests to start switched link activation sequences. There are many
reasons for delays in this step (such as waiting for operator intervention
to answer a message or dial a switched connection). This step can also
cause a controller description to be automatically created by the oper-
ating system (which can also cause a delay).
Device selection This processing measures the number of times that the T2
station IOM task is requested to select a device description. This step
can lead to the automatic creation and/or vary on of device descriptions.

8-20 Performance Tools V4R2

 The measurements for automatic creation and/or vary on of device
descriptions can be found in the QAPMSNA performance monitor file.

APPC Protocol
For APPC, there are two types of transactions for which the Performance Tools
collect sample data: inbound and outbound.

An outbound transaction begins when the request is issued and ends when the
complete response is received. An inbound transaction begins when a request is
received and ends when a response is sent.

The transaction timings provide a picture of how much time is spent in processing a
transaction on the local and remote systems.

System A Remote System B

Program Program
6 3

5 1 2 4
Comm line

Comm IOP Comm line Comm IOP

RV2S089-3

Figure 8-3. Using APPC Protocol Communications

From System A's Perspective

System A is sending data to system B and is expecting a response from system B.
Figure 8-3 shows the steps in system A’s outbound transaction.
1. System A puts, or transmits, data
2. System B gets, or receives, the data from system A
3. A program on system B processes the data and produces new data in
response
4. System B puts the new data, which is its response to system A
5. System A receives the data, which is the response from system B
6. The program on system A processes data, determines the next request, and
returns to step 1

The following ratios of fields found in the QAPMJOBS file are effective performance
indicators:
JBPUTA/JBPUTN The average number of bytes per put operation. Larger values
indicate greater efficiency because fewer put operations are necessary.
JBPUTA/JBRTI The average number of bytes buffered per transmit (request). Put
operations made by system A go into a communications buffer. The
system transmits the contents of the buffer when it fills up with oper-

Chapter 8. Transaction Boundaries—Manager Feature 8-21

 ations requiring data to flow to the remote system. The higher this
number is, the better the buffer is utilized. When the buffer is used more
efficiently, the lower communications layers are called less often and the
result is better performance.

From System B's Perspective

System B is receiving information from system A and will be sending a response to
system A. In Figure 8-3 on page 8-21, steps 2, 3, and 4 make up system B's
inbound transaction. This is a subset of system A's outbound transaction.

2. System B gets, or receives, the data from system A

3. A program on system B processes the data and produces new data in
response
4. System B puts the new data, which is its response to system A

Conversely, steps 4, 5, 6, 1, and 2 can be considered system B's outbound trans-

action. Steps 5, 6, and 1 make up system A's inbound transaction, which is a
subset of system B's outbound transaction.

The following ratios of fields found in the QAPMJOBS file are effective performance
indicators:
JBGETA/JBGETN The average number of bytes per get operation. Larger values
indicate greater efficiency because fewer get operations are necessary.
JBGETA/JBRRI The average number of bytes made available to the program in a
buffer. The receiving system buffers data received until the buffer fills up
or until an operation (on the transmitting system) requires the immediate
delivery of data to the receiving program.

Notes:
1. A series of multiple put operations can be created by system A. At the end of
these put operations, system A effectively passes control to system B through a
change direction (CD) operation. System B can complete its first get operation
when a buffer of data arrives from system A. It can process subsequent put
operations from system A as it receives them. However, it cannot send any
data back to system A until it receives the change direction (CD) indication.
2. System A's put operations can overlap system B's get operations.

APPC Performance Notes

A complete end-to-end analysis of APPC performance includes the following
elements:
Ÿ Communications time
Ÿ IOP time on system A (outgoing and incoming)
Ÿ Line time (from system A to system B, and from system B to system A)
Ÿ IOP time on system B (incoming and outgoing)
Ÿ Processing time on system B

Knowledge of the application design, along with performance data from both
system A and system B, allows you to analyze the application's performance. Using
the outbound time (in the JBPGIL field in the QAPMJOBS file) on the local system

8-22 Performance Tools V4R2

 and the inbound time (in the JBGGIL field in the QAPMJOBS file) on the remote
system, the amount of time the local system spent waiting for the remote system to
respond to its request can be determined. With the addition of information such as
the number of input and output operations plus the amount of data sent and
received, average transaction times can be determined. The line speed can be
used to determine how much the line slows processing and the effect of changing
this. The processing unit in the application can be used on a per transaction basis
(where communications activity is continuous) to determine the effects of model
upgrades.

The Work Management book contains complete information on all the APPC infor-
mation collected by the performance monitor.

Performance Measurement and SNADS

In general, SNADS sample data does not provide any direct performance tuning
capabilities. There are no distribution parameters that can be adjusted to affect
SNADS performance (with a minor exception for distribution queue attributes and
SNADS senders). However, the customer can do normal system job tuning and
measure the results in distribution throughput using the sample data.

Performance tuning for SNADS jobs may involve adjusting the job priority or other
attributes contained in the job class or job description. A description of each job
and the characteristics important for performance analysis and tuning is provided in
Work Management. Unless otherwise noted in the detailed sections, SNADS jobs
have the following common characteristics:
Ÿ All are submitted using the QSNADS job description.
Ÿ A separate routing entry exists in the QSNADS subsystem for each type of
SNADS job. This allows the customer to identify different job classes (priority)
for each. The default is class QSNADS, which has a run priority of 40.
Ÿ All run under the QSNADS or QGATE user profile.
Ÿ All internal distribution objects (those not visible to the user) created by SNADS
are owned by the QSNADS user profile. This identifies how much system
storage is being used by distribution activities.

SNADS Transaction
A SNADS transaction and a distribution within a SNADS job generally have a one-
to-one relationship. A SNADS transaction is the processing done by a SNADS job
as it handles a distribution. Each distribution processed is considered a transaction,
including both distributions processed successfully and distributions processed with
errors.

The following functions in a SNADS transaction are described in this section:

Ÿ Router
Ÿ Receiver
Ÿ Sender
Ÿ Gateway senders

Chapter 8. Transaction Boundaries—Manager Feature 8-23

 Notes:
1. Distribution processing that ends with errors may result in another attempt of
the same distribution. Thus, a single distribution may cause the transaction
count to be incremented multiple times. For example, the line drops during a
transmission. Each send attempt counts as a transaction even though the same
distribution is being processed each time. Each transaction (ending with an
error) accounts for resources expended in the distribution attempt.
2. The number of transactions ending with errors is included in the error count
provided in the sample data.

Table 8-1. SNADS Transaction Types

Transaction Type SNADS Function
'01' SNADS Router
'02' SNADS Receiver
'03' SNADS Sender
'04' SystemView Distribution Services (SVDS) Receiver
'05' SVDS Sender

SNADS Gateway Senders

'08' SNADS DLS Gate (Document Library Services)

'09' SNADS Gate (VM/MVS Bridge, SMTP, X.400)

SNADS Router
The QSNADS router function is the heart of SNADS. All distributions that flow
through SNADS are processed by the router. It uses the system directory and dis-
tribution services configuration to determine what queue or queues to put the dis-
tribution on.

A distribution with multiple destinations may have a distribution copy placed on a

local delivery queue and multiple distribution queues. This is called fan-out. Some
destinations may be routed successfully while others result in router errors.

A router transaction begins when the router finds a distribution on its queue. The
transactions ends when the distribution has been placed on all applicable queues
and removed from the router queue.

The SNADS router function has the following characteristics:

Ÿ The SNADS router function is a job that runs in the QSNADS subsystem and is
started by the SNADS startup job (QZDSTART).
Ÿ The job name is QROUTER.
Ÿ The job's user profile and job description are QSNADS.
Ÿ The subsystem routing entry compare value is 'QROUTER'.
Ÿ The router job should remain active as long as the QSNADS subsystem is
active. If the router ends or is canceled, the subsystem must be stopped and
restarted to start the router.

8-24 Performance Tools V4R2

 Ÿ The router function can support only one router job.

SNADS Receiver
The SNADS receiver function is a job that runs in response to a SNADS remote
sender opening a session and doing an APPC evoke operation. The SNADS
receiver manages the receive side communications protocol for the SNADS conver-
sation.

A receiver transaction begins when the receiver receives distribution data from the
sender. The distribution data is separated and stored in an internal control block. A
file server object is created if the distribution carried one. The distribution is put on
an internal queue for the SNADS router to process. The receiver then logs and
sends confirmation to the sender function. The transaction ends when the confir-
mation request is complete or the job ends for any reason (for example, a commu-
nications error).

A SNADS receiver job has the following characteristics:

Ÿ It runs in the subsystem configured for the communications device that was
started by a remote sender. The default is QCMN subsystem.
Ÿ The job name is the same as the communications device name.
Ÿ The job's user profile, job description, and so on are determined by the subsys-
tem's communications entry and routing entries. These normally default to the
QSNADS user profile, QSNADS job description, and PGMEVOKE subsystem
routing entry.
Ÿ A receiver job runs until one of the following occurs:
– The evoking sender no longer has data to send.
– An error is detected by either the sender or receiver.
– The session ends abnormally (for example, the line fails).
Ÿ A receiver may be started repetitively over time by a job on the sending system
(probably the same sender job).
Ÿ SNADS imposes no limits on the number of receiver jobs that can be active at
the same time. Multiple receivers can be active for the same device.
Ÿ The QSNADS subsystem does not have to be active. Receivers can queue dis-
tributions for the router job regardless of the state of the router or QSNADS
subsystem.

SNADS Sender
The SNADS sender function manages the send side communications protocol for
the SNADS conversation. It starts the remote receiver and sends any available dis-
tributions (queued on its distribution queue) to the remote system. SNADS senders
service *SNADS type distribution queues.

The sender is sensitive to the following:

Ÿ The state (held or waiting) of the distribution queue.
Ÿ The state (held or waiting) of the queue entries.
Ÿ The send conditions configured for the queue.

Chapter 8. Transaction Boundaries—Manager Feature 8-25

 The sender does not establish a communications session or send any distribution
unless its distribution-queue conditions or states allow it.

When send conditions are met or they end, a certain amount of overhead is
required to establish or end communications. This is not included in the sender
transaction (resource use time); it is, however, included in the overall job statistics.
If errors occur during this activity, the sample data error count is incremented along
with the active transition count but no other transaction data or counts will change.

A sender transaction begins when the sender dequeues the next distribution to be
sent (send conditions were previously met and a session is active). Distribution
data is put into code and sent to the receiver. If a file server object is present, that
data is read and sent with the distribution. The sender waits for the receiver to
confirm the distribution, at which time it is logged and removed from the distribution
queue. This ends the transaction. The transaction can also be ended by any error
that is detected during processing.

The SNADS sender function has the following characteristics:

Ÿ The SNADS sender function is a job that runs in the QSNADS subsystem and
is started by the SNADS startup job (QZDSTART). One job is started for each
distribution queue (type *SNADS) configured. Sender jobs may also be started
by configuration (CFGDSTSRV) as distribution queues are added, and by oper-
ations (WRKDSTQ) if the operator starts a send operation.
Ÿ The job name is the same as the remote location name configured with the
distribution queue.
Ÿ The job's user profile and job description are QSNADS. The subsystem routing
entry compare value is 'QSENDER'.
Ÿ The sender job should remain active as long as the QSNADS subsystem is
active, if no errors are detected. Sender jobs for queues with no time or depth
specified end as soon as the queue is empty. If the sender is detecting errors
and the retry count configured for the queue is exceeded, the sender job ends.
The sender can be restarted by selecting the Send queue option on the
WRKDSTQ display or issuing the SNDDSTQ command.
Ÿ There can be as many sender jobs as there are distribution queues.
Ÿ The sender function can support only one sender job per distribution queue.
(Multiple sender jobs may sometimes be started for a particular queue but they
will eventually cancel each other until only one is left.)
Ÿ SNADS senders can be started (evoked as an APPC application) by remote
SNADS receivers. Although this function is supported by AS/400 senders, there
is no known SNADS implementation where a receiver will start a sender.

SVDS Receiver
The SVDS receiver function is a job that runs in response to an SVDS remote
sender opening a session and doing an APPC evoke operation. The SVDS receiver
manages the receive side communications protocol for the SVDS conversation.

A receiver transaction begins when the receiver receives distribution data from the
sender. The distribution data is separated and stored in an internal control block. A
file server object is created if the distribution carried one. The distribution is put on
an internal queue for the SNADS router to process. The receiver puts a completion
report message unit on a queue. This ends the transaction. The transaction can

8-26 Performance Tools V4R2

also be ended by an error detected during processing. The receiver will send the
completion report message unit to the sender after the sender sends a change
direction (CD) indication. The receiver processes any other transactions from the
sender. After the sender sends a CD indication, the receiver sends the completion
report message unit.

An SVDS receiver job has the following characteristics:

Ÿ It runs in the subsystem configured for the communications device that was
started by a remote sender. The default is QCMN subsystem.
Ÿ The job name is the same as the communications device name.
Ÿ The job's user profile, job description, and so on are determined by the subsys-
tem's communications entry and routing entries. These normally default to the
QGATE user profile, QSNADS job description, and PGMEVOKE subsystem
routing entry.
Ÿ A receiver job runs until one of the following occurs:
– The evoking sender no longer has data to send.
– An error is detected by either the sender or receiver.
– The session ends abnormally (for example, the line fails).
Ÿ A receiver may be started repetitively over time by a job on the sending system
(probably the same sender job).
Ÿ There can be only one receiver job active per connection.
Ÿ The QSNADS subsystem does not have to be active. Receivers can queue dis-
tributions for the router job regardless of the state of the router or QSNADS
subsystem.

SVDS Sender
The SVDS sender function manages the send side communications protocol for the
SVDS conversation. It starts the remote receiver and sends any available distrib-
utions (queued on its distribution queue) to the remote system. SVDS senders
service *SVDS type distribution queues.

The sender is sensitive to the following:

Ÿ The state (held or ready) of the distribution queue.
Ÿ The state (held, ready, pending, or suspended) of the queue entries.
Ÿ The send conditions configured for the queue.

The sender does not establish a communications session or send any distribution
unless its distribution-queue conditions or states allow it.

When send conditions are met or they end, a certain amount of overhead is
required to establish or end communications. This is not included in the sender
transaction (resource use time); it is, however, included in the overall job statistics.
If errors occur during this activity, the sample data error count is increased along
with the active transition count. No other transaction data or counts will change.

A sender transaction begins when the sender dequeues the next distribution to be
sent (send conditions were previously met and a session is active). Distribution

Chapter 8. Transaction Boundaries—Manager Feature 8-27

 data is put into code and sent to the receiver. If a file server object is present, that
data is read and sent with the distribution.

The SVDS sender function has the following characteristics:

Ÿ The SVDS sender function is a job that runs in the QSNADS subsystem and is
started by the SNADS startup job (QZDSTART). One job is started for each
distribution queue (type *SVDS) configured. Sender jobs may also be started
by configuration (CFGDSTSRV) as distribution queues are added, and by oper-
ations (WRKDSTQ) if the operator starts a send operation.
Ÿ The job name is the same as the remote location name configured with the
distribution queue.
Ÿ The job's user profile is QGATE. The job description is QSNADS. The sub-
system routing entry compare value is 'QSVDSSND'.
Ÿ The sender job should remain active as long as the QSNADS subsystem is
active, if no errors are detected. Sender jobs for queues with no time or depth
specified end as soon as the queue is empty. If the sender is detecting errors
and the retry count configured for the queue is exceeded, the sender job is
suspended. The sender can be restarted by selecting the Send queue option
on the WRKDSTQ display or issuing the SNDDSTQ command.
Ÿ There can be as many sender jobs as there are distribution queues.
Ÿ The sender function can support only one sender job per distribution queue.
(Multiple sender jobs may sometimes be started for a particular queue but they
will eventually cancel each other until only one is left.)
Ÿ SVDS senders can be started (evoked as an APPC application) by remote
SVDS receivers. Although this function is supported by AS/400 senders, there
is no known SNADS implementation where a receiver will start a sender.

SNADS Gateway Senders (DLS Gate and VM/MVS Bridge)

Gateway senders are not a function of SNADS architecture. AS/400 SNADS
support provides distribution queuing and scheduling support for other distribution
functions. This support is provided through distribution queues (queue types *DLS
and *RPDS) and the SNADS gateway sender function.

Gateway senders are similar in every respect to SNADS senders except that
SNADS does not handle any communications nor does it matter if the distribution
ever leaves the local system. Based on the same queuing controls as SNADS
senders, distributions are handed over to the appropriate bridge function. When the
bridge function confirms that it has successfully received (or processed) the distrib-
ution, the distribution is removed from the SNADS queue.

The transaction begins when it is time to send and a distribution is found on the
queue. The distribution data is put into code for the bridge function along with any
file server object. The gate sender waits for a response from the bridge indicating
the distribution was sent; then the distribution is logged and removed from the
queue. This ends the transaction. Any error detected by the gateway sender or an
error response from the bridge would also end the transaction.

Gateway senders have the same characteristics as the SNADS senders except:
Ÿ The job's user profile is QGATE.
Ÿ The subsystem routing entry compare value is 'QGATEWAY'.

8-28 Performance Tools V4R2

 Ÿ The bridge function may or may not completely process the distribution under
the gateway sender job. All current implementations process the distribution in
other jobs. Therefore, the sample data only reflects the resource required to
hand over the distribution. One possible exception is the resource use time.
This may reflect total time, depending on when the bridge function acknowl-
edges receipt of the distribution.

Table 8-2. Sample Data for Each SNADS Function

SNADS or SNADS or SNADS
Data Field SNADS SVDS SVDS SNADS RPDS
Description Field Name Router Receiver Sender DLS Gate Gate
Transaction SNNTR X X X X X
count
Resource use SNRUT X X X X X
time
FSO count SNFSO X X X X X
FSO byte count SNFSOB X X X X X
Transaction time SNTRT X X X X X
Error count SNERR X X X X X
Active transi- SNATN X X X X X
tions count
Recipient count SNNRC X X X X X
Fan-out count SNFOC X – – – –
Local distribution SNLDC X – – – –
count

SNADS Sample Data

This section describes the sample data provided by SNADS. The data collected is
the same for all SNADS jobs (that is, the sample data format does not change).
However, not all entries apply to every SNADS function (sample data subtype).
Data that does not apply to a function will be zero when the data is written.

Table 8-2 summarizes the sample data supported for each SNADS function.

Sample Data Interpretation

The primary purpose of the SNADS specific sample data is performance planning.
This data provides statistics on what SNADS activity is taking place over time. It
indicates the amount, size, and location of the distribution load on the system.

This sample data does not include resources used for local distribution (from a local
user to a local user). SNADS involvement is limited to asynchronous remote distrib-
ution. This includes remote systems sending mail to the local system, the local sys-
tem's role as an intermediate node, and local distribution to a remote system.

The following are various SNADS jobs described by transaction type:

Ÿ The router job (type X'01') indicates the total amount of mail being handled by
the system. This reflects distributions from receivers, distributions originated

Chapter 8. Transaction Boundaries—Manager Feature 8-29

 locally, distributions arriving from one of the SNADS bridge interfaces, as well
as DSNX-PC activity. Assuming there is not substantial bridge activity, the
number of distributions originating locally is the difference between the distrib-
utions routed and the distributions received.
Ÿ SNADS receiver jobs (type X'02') and SVDS receiver jobs (type X'04') indi-
cate the amount of mail arriving on the local system from remote SNADS
sources. Specific receiver jobs indicate the amount arriving from the associated
location. Receiver jobs must be processed in an aggregate because the
sending system evokes a receiver job whenever it has something to send. This
causes receiver jobs to start and stop often.
Ÿ SNADS sender jobs (type X'03'), SNADS gateway sender jobs (types X'08'
and X'09'), and SVDS sender jobs (type X'05') represent distributions leaving
the system. These may originate from all the same sources as described above
for the router. Because there may be multiple paths off a system (connections
to multiple destinations), a single distribution processed by the router may be
copied to multiple distribution queues and therefore represent a sender trans-
action in multiple sender jobs.

For SNADS receivers, routers, and senders, the job data reflects the real proc-
essing time, and for receivers and senders it also reflects the APPN/APPC
resources used. Gateway senders are a little different in that the data is only
handed over to a bridge function; there may be additional processing in other jobs.
No communications take place in a gateway sender.
Note: Although SVDS senders are defined with user profile QGATE, they do use
communications.

SNADS Sample Data Field Descriptions

Transaction count (SNNTR): Under normal conditions, the transaction count indi-
cates how many distributions have been processed by the respective jobs. In the
case of the router, it also indicates how many SNADS distributions are flowing
through the system.

Transaction count is not an absolute distribution count:

Ÿ A router error indicates one or more recipients failed. If no recipients were
routed successfully, the distribution goes no further. If at least one recipient
was valid, the distribution is placed on one or more queues.
Ÿ Receiver and sender errors usually represent distribution attempts that failed
rather than distributions that failed (for example, when the line goes down); at
some future time, the distribution will be successful. Therefore, when this type
of error has occurred, the transaction count is incremented once when the dis-
tribution attempt is successful and one or more times when it fails. Note,
however, that certain irrecoverable errors can occur and result in the deletion of
the distribution.
Ÿ Distributions can be deleted from queues by operator action. This can occur
during distribution or between distribution attempts.
Ÿ Transaction count is not updated when an error occurs prior to the start of the
transaction—even though the error count may be updated.

Resource Use Time (SNRUT): This indicates how long it took the job to process
the transaction. It is a function of system loading, relative job priorities, and commu-
nications line speed (receivers/senders). Another important cause is the size of the

8-30 Performance Tools V4R2

distribution. Distributions that carry large amounts of data (see FSO count) take
longer to send and receive.

Longer resource use times (especially for senders) also mean longer transaction
times for subsequent queued distributions.

Size (FSO) (SNTRT): Some distributions are very small (for example, messages)
and do not require much communications resource to transfer. Other distributions
can carry data objects or documents. In addition to the obvious effect on communi-
cations, these distributions require added resource and processing time to be
stored on or copied from the local system (only one copy is made).

The amount of data being carried increases all resources and measurements asso-
ciated with the job.

The FSO count (SNFSO) provides a comparison between the number of trans-
actions with and without file server objects. Message distributions do not file server
objects associated with them. The FSO byte count (SNFSOB) also indicates how
much data is being moved by the distributions carrying FSOs.

Errors: The error count should normally be very low or 0. Some router errors may
be expected and reflect an user ID that is not valid or a system that was not
entered in the routing table. Some sender errors may be expected if the remote
system is down or there is a line problem.

In this case, the error count (SNERR) and active transitions count (SNATN)
increase without affecting other transaction counts and data.

If line performance is a problem, error rates for senders and receivers may indicate
that the line is not staying up long enough to complete sending of a distribution,
causing distributions to be sent multiple times.

Recipients: The recipient count (SNNRC) indicates how many users are in the
destination list of the distribution. These may be individual users or names of dis-
tribution lists that expand at the destination systems. The number of recipients has
a major effect on the router but little effect on senders and receivers.

Distributions with multiple recipients may go to multiple destination systems through

different distribution queues as well as to the local system. One distribution copy is
placed on each queue regardless of how many recipients belong to that queue. The
router fan-out count (SNFOC) and local distribution count (SNLDC) indicate
where distributions are going (remote/local) and to how many different queues.

Transitions and Queuing Time: The active transitions count indicates how often
the job was waiting for a distribution to process (provided other controls did not
prevent activity) or how often a sequence of one or more distributions was proc-
essed. Queuing time is a measure of distribution delay. Queuing time is the differ-
ence between transaction time and resource use time.
Ÿ For the router, job transitions are not expensive. A high transition rate indicates
that the router is moving distributions quickly. A low transition rate combined
with longer queuing times may indicate that the router's job priority is too low
for the distribution rate.

Chapter 8. Transaction Boundaries—Manager Feature 8-31

 Ÿ Senders do not send unless queue conditions are met. These include: queue is
ready, time is within from/to time, queue depth is reached, and distributions are
in the queue. When the conditions are met, the sender does the following:
– Passes to an active state (queue state changes from Waiting to Sending).
– Tries to establish a session.
– Tries to start the remote receiver.
Because this is substantial additional distribution overhead (and the most error
prone), the active transitions count was provided.
If the transition count represents a high percentage of sender transactions, and
performance is vital, an increase to the queue depth should be considered.
Queuing time can be used to balance the distribution delay on the send queue
as opposed to the number of transitions.
A high sender transition rate accompanied by high error counts might also indi-
cate that the sender is experiencing difficulty establishing communications and
is in a wait-retry recovery loop.
Ÿ On a target system, a receiver job is started every time the sender job makes
an active transition. This is additional overhead that is also reduced by reducing
the remote sender transitions as described above.
The transition count for receiver jobs parallels the transition count for the asso-
ciated remote sender. A high transition count relative to the number of distrib-
utions received from a remote location may indicate a need to make some
remote sender adjustments or consider using prestart jobs for SNADS
receivers.

SNADS Performance Notes

1. The run priority of SNADS jobs can be changed by changing the class of the
corresponding subsystem routing entry.
Ÿ The router's priority can be raised to increase throughput when there are
periods of heavy interactive activity using SNADS. System performance
may be satisfactory even if the router has a priority higher than interactive.
However, if a system problem occurs (for example, a loop), the router could
take over the system.
Ÿ The receiver's priority can be raised by adding a routing entry with a
compare value of 'QZDRCVR' at position 37 of the routing data.
2. When distributions continually arrive on a sender queue, but at a rate slower
than required to send, system resource will be wasted, constantly starting and
stopping sessions on both local and remote systems as well as starting jobs on
the remote system.
Ÿ The queue depth can be used to allow some number of distributions into
the queue before sending begins. This also has the benefit of allowing
additional distributions to arrive on the queue while the queue is being
emptied. However, the distribution is delayed until enough distributions
accumulate to satisfy the send conditions. Distribution rates, queuing times,
and queue active (send) transitions are available in the sample data to
measure this.

8-32 Performance Tools V4R2

 Ÿ Prestart jobs can be used on the remote system to reduce receiver job
startup overhead.
3. Activity to or from a particular remote location can be determined by looking at
the data for the senders and/or receivers associated with that remote location.
The job name is used to select these jobs.
Each sender is named by the remote location name specified in the SNADS
configuration and the corresponding device. Each receiver job name is the
same as the device carrying the conversation. If device names are the same as
the remote location name (usually true), all send and receive activity to a partic-
ular remote location can be observed using the remote location as the job
name.

OS/400 File Server

Transaction information is collected for two types of file server shared folders trans-
actions. Both transaction types are handled within the Licensed Internal Code. The
transaction types are the following:
Ÿ Requests from personal computers; the transaction type is *TNS (Transaction).
Ÿ Replies sent to the personal computer through T2; the transaction type is
*QUEUE (Measuring queuing times).

For the first transaction type (request from a personal computer), both the time the
request is received and the time the request is finished are logged. For the
Licensed Internal Code requests, the times should be very short. Locks/Unlocks
typically are only one or two milliseconds. Reads/Writes depend on how much
input/output needs to be done. Change End of File, Force Buffers, and Resets of
files are also handled.

The OS/400 program handles the following requests:

Ÿ Create
Ÿ Delete
Ÿ Open
Ÿ Close
Ÿ Directory (List file attributes)
Ÿ Make directory
Ÿ Remove directory

These requests are not logged for transaction data.

Personal computer functions like the copy or type function are classified into mul-
tiple requests (usually list file attributes, open file, read/write, and close). Only the
times for read and write requests are logged.

The second transaction type described (replies sent to the personal computer) logs
the amount of time it takes for T2 to respond to the file server stating that the reply
(from the file server to the personal computer) has been sent. This is done for com-
mands handled by both the OS/400 program and the Licensed Internal code. Also,
a single command from the personal computer, like a read request, could result in

Chapter 8. Transaction Boundaries—Manager Feature 8-33

 multiple queuing operations. These times should also be short. Figure 8-4 on
page 8-34 shows this transaction type.

Personal Computer T2 AS/400 System

PC Application 2 Server
1
3

8 5 4

6 7
RV2S072-3

Figure 8-4. T2 transaction type

1. PC application sends request to T2

2. T2 sends the request to Licensed Internal Code server
3. Server logs the following
a. Start of transaction
b. End of transaction
c. Start of queue transaction
4. Licensed Internal Code server sends reply to T2
5. T2 sends the reply to PC application
6. T2 sends acknowledgment to Licensed Internal Code server
7. Licensed Internal Code server logs the end-of-queue transaction
8. PC application receives a reply

The transaction times start after the server gets the request and end before a reply
is sent.

The total transaction time (in seconds) is stored in the JBRSP field in the
QAPMJOBS file. The number of transactions (5250 only) is stored in the JBNTR
field in the QAPMJOBS file. These fields are also updated by display I/O trans-
actions and by pass-through transactions.

OS/400 file server jobs run in the QSERVER subsystem.

Pass-Through Transactions

Source System Target System

Enter 3 4
1 2

View 4
Display 6 5

RV2S071-1

Figure 8-5. Source Pass-through Transaction Path

8-34 Performance Tools V4R2

 Figure 8-5 shows the flow of data for a pass-through transaction. The following
occurs:
1 A user presses the Enter key, sending a request for data.
2 The request for data is then sent to the target system.
3 The transaction data leaves the source system.
4 The transaction data is on the network.
5 The transaction data is sent from the target system to the source
system.
6 Data appears on the display.
1→2 Data is in the source system Pass-through is processing data
2→3 Service time
3→5 Display wait time
5→6 Data is in the source system
1→6 Transaction

Pass-Through Performance Notes

The target program on a session time includes only trace points (no sample data).
Here are some useful formulas:
Display wait time (3→5) The time from when the source system sends a request
to when it receives the data from the target system.
Service time (2→3) The time from when the source system sends a request for
data to the target system until the requests are on the network.
Transaction time (1→6) The time from when a user presses the Enter key until a
new display appears.
The total transaction time (in seconds) is stored in the JBRSP field in
the QAPMJOBS file. The number of transactions (5250 only) is stored in
the JBNTR field in the QAPMJOBS file. These fields are also updated
by display I/O transactions and by Client Access shared folders trans-
actions.

Data Queue Transactions

Data queues provide a means for one job to start a work activity in another server
job. A data queue transaction identifies and provides a means of measurement for
this server work activity with the following boundaries:

Program A Program B

RV2S091-1
Data Queue

Figure 8-6. Data Queue Transaction

Chapter 8. Transaction Boundaries—Manager Feature 8-35

 1. Program A sends data to the data queue. This is the start of the application
input queuing time.
2. Program B dequeues the data. This ends the application input queuing time.

The total application queuing time (in hundredths of a second) is stored in the
JBAIQT field in the QAPMJOBS file. The number of application queuing trans-
actions is stored in the JBNAIQ field in the QAPMJOBS file. These fields are
also updated by display I/O transactions. This also starts the resource utiliza-
tion time.
3. Program B dequeues the next data. This ends the resource utilization time.

The total resource usage time (in seconds) is stored in the JBRUT field in the
QAPMJOBS file. The number of resource usage transactions is stored in the
JBNRU field in the QAPMJOBS file. These fields are also updated by display
I/O transactions.

8-36 Performance Tools V4R2

Chapter 9. Performance Graphics
This chapter describes the functions that allow you to work with performance data
in a graphical format. The performance data is collected using the Start Perform-
ance Monitor (STRPFRMON) command. The graphs can be displayed interactively,
printed, plotted, or saved to a graphics data format (GDF) file for use by other utili-
ties, such as the Business Graphics Utility (BGU).
Note: This chapter does not refer to Capacity Planning graphics. For information
about using the graphics feature of the capacity planner, see the BEST/1
Capacity Planning Tool book.

Summary—Manager Feature
Two distinct types of graphs can be displayed: performance graphs and historical
graphs. Performance graphs use the performance data collected from a single run
of the performance monitor. Performance graphs are useful for singling out jobs
that are performing poorly or evaluating the activities performed by a user or class
of users on the system during a specified period.

Historical graphs use performance data collected from several runs of the perform-
ance monitor. Historical data is the summary of the performance data created by
the STRPFRMON command. The Create Historical Data (CRTHSTDTA) command
is used to summarize the performance data for use by the historical graphs. Histor-
ical graphs are used to show how the performance of a system has changed over
time.

1. Create a graph format using the Create Graph Format (CRTGPHFMT)

command. (Graph formats are reusable.)
2. Collect performance data using the STRPFRMON command.
3. Display the graph using the Display Performance Graph (DSPPFRGPH)
command.

Use the following steps to display a historical graph:

1. Create a graph format using the CRTGPHFMT command. (Graph formats are
reusable.)
2. Collect performance data using the STRPFRMON command.
3. Create the historical data using the CRTHSTDTA command.
4. Display the graph using the Display Historical Graph (DSPHSTGPH) command.

When you select option 9 (Performance graphics) on the IBM Performance Tools
menu, the Performance Tools Graphics menu appears.

 Copyright IBM Corp. 1998 9-1

 à ð
PERFORMG Performance Tools Graphics
System: ABSYSTEM
Select one of the following:

1. Work with graph formats and packages

2. Work with historical data
3. Display graphs and packages

7ð. Related commands

Selection or command
===> __________________________________________________________________
_______________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant
F16=System main menu

á ñ
You can also reach this menu by typing the Start Performance Graphics
(STRPFRG) command on the command line of any display. From this menu, you
can work with performance data in a graphical format.

Work with Graph Formats and Packages—Manager Feature

Graph formats are templates or outlines used by the DSPPFRGPH and the
DSPHSTGPH commands to display graphs in a user-defined format. Table 9-1
shows the 15 predefined graph formats included in QPFRDATA, the IBM-supplied
performance library.

Table 9-1 (Page 1 of 2). QPFRDATA Graph Formats

Graph Format Name Description
QIBMASYNC Asynchronous disk I/O per second against
time
QIBMCMNIOP Communications IOP use against time
QIBMCMNLIN Maximum communications line use against
time
QIBMCPUPTY Processor unit use of jobs with priorities
0-19, 20-39, 40-59, 60-79, and 80-99
against time
QIBMCPUTYP Processor unit use of batch, interactive, and
system jobs against time
QIBMDSKARM Disk arm use against time
QIBMDSKIOP Disk IOP use against time
QIBMLWSIOP Local work station IOP use against time
QIBMMFCIOP Multifunction communications IOP use
against time
QIBMMFDIOP Multifunction disk IOP use against time
QIBMDSKOCC Percentage of disk occupied against time
QIBMRSP Interactive response time against time
QIBMTOTDSK Total disk I/O per second against time

9-2 Performance Tools V4R2

Table 9-1 (Page 2 of 2). QPFRDATA Graph Formats
Graph Format Name Description
QIBMTNS Transactions per hour against time
QIBMSYNC Synchronous disk I/O per second against
time

Graph packages allow you to group several graph formats into a single entity. This
is useful for printing, displaying, or plotting a number of graphs at once. Instead of
having to issue several DSPPFRGPH or DSPHSTGPH commands to print several
graphs, you can use the package name (one command) to print all of the graphs in
the graph package. Also included in QPFRDATA is a predefined graph package,
QIBMPKG, which contains the 15 IBM standard graph formats.

If you select option 1 (Work with graph formats and packages) on the Performance
Tools Graphics menu, the Work with Graph Formats and Packages display
appears.

à ð
Work with Graph Formats and Packages

Library . . . . . . . . . . . QPFRDATA

Type options, press Enter.

1=Create graph format 2=Change 3=Copy 4=Delete
5=Display sample graph 6=Create package 8=Display package contents

Format/
Option Package Type Text
_ __________ _________________________________________
_ PACKAGE1 PACKAGE Graph Package containing format w/ func. areas
_ QIBMPKG PACKAGE IBM Graph Package
_ FORMAT1 FORMAT CPU Utilization vs. Time-Functional Areas
_ NWCTEST FORMAT
_ QIBMASYNC FORMAT Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP FORMAT Communications IOP Utilization vs. Time
_ QIBMCPUPTY FORMAT CPU Utilization vs. Time (Priority)
_ QIBMCPUTYP FORMAT CPU Utilization vs. Time (Job Type)
_ QIBMDSKARM FORMAT Disk Arm Utilization vs. Time
More...

F3=Exit F5=Refresh F12=Cancel F15=Sort by format F16=Sort by text

á ñ
This display shows you the graph formats and graph packages that exist in the
library specified in the Library field. The graph format or graph package name, a
format or package indicator, and a text description appear on the display. If you
cannot find the format or package you want to work with, use the appropriate func-
tion key to sort the formats and packages. You can sort them by name, type, or
text description. When you find the graph format or package you want to work with,
select the function you want to perform by typing the appropriate option in the
Option field and pressing the Enter key.

If you are searching for a graph format or graph package located in a library that is
different from the one currently listed in the Library field at the top of the display,
type a new library name in the field and press the Enter key. A list of the graph
formats and graph packages available in the library you specified appears. You can
then select one of them to work with.

Chapter 9. Performance Graphics 9-3

Create Graph Format—Manager Feature
To create a new graph format, type a 1 (Create graph format), the graph format
name, and the description on the first line under the Option, Format/Package, and
Description columns, and press the Enter key. The CRTGPHFMT command prompt
appears.

Specify how your graphs are displayed by selecting from the following options:
Ÿ Titles
Ÿ X-axis data
Ÿ Y-axis data
Ÿ Data type
Ÿ Individual line breakdown
Ÿ Graph type

Graph Types—Manager Feature

The graph types available are:
Ÿ Line
Ÿ Scatter plot
Ÿ Surface
Ÿ Floating bar
Ÿ Composite bar

Line Graphs
Use line graphs to show change occurring over time. Line graphs can represent
increases, decreases, trends, and general fluctuations of quality.

CPU Monday’s Performance Data

Utilization CPU Utilization vs. Time (Job Type)

80
70
60
50
40
30
20
10
0
08:00 10:00 12:00 14:00 16:00 18:00
Time
Press Enter to continue. Batch
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot System
RV2S084-0

Figure 9-1. Line Graph: Data Represented as Lines

Each plotted point is shown by a marker; the plotted points are connected to form a
continuous line. Each line is assigned a different color. If lines overlap, the color of
the last legend entry at that point is displayed.

9-4 Performance Tools V4R2

Scatter Plots
Scatter plots are similar to line graphs, except that the lines that connect the data
points are not drawn.

CPU Monday’s Performance Data

Utilization CPU Utilization vs. Transactions per Hour
x
80
70 + x

60 x
x
50 + x x
+
40 x
x x x x
30 + x
+ x x x x
20 + x xx x xx x x
x xx
x x x x x
10 +x x
xx
x x
x
0x
0 200 600 1000 1400 1800 2200 2600 3000
Transactions per Hour
Press Enter to continue. + Batch
F3=Exit F6=Print F9=Overlay x Interact
F12=Cancel F16=Save GDF F17=Plot System
RV2S078-0

Figure 9-2. Scatter Plot: Data Represented as Markers

Surface Graphs
Use surface graphs like line graphs to show changes occurring over time. Surface
graphs emphasize volume by shading the area between the lines and the X-axis if
you specify Y (yes) for the area fill option.
Note: If you do not use the area fill option in your surface graph, your graph will
be a cumulative line graph. If there is a legend entry with a value of zero to
plot, its line covers the line plotted previously because there is no change to
plot. Although shading requires more time to display or plot than simply
drawing the lines, the area fill option may show more clearly which legend
entries represent the different areas, particularly in cases where a line of
one color may cover another.

CPU CPU Utilization vs. Time (Job Type)

Utilization
55
50
45
40
35
30
25
20
15
10
5
0
08:00 08:30 09:00 09:30 10:00 10:30
Time
Press Enter to continue. Batch
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot System

RV2S081-0

Figure 9-3. Surface Graph: Data Represented as Shaded Regions

Chapter 9. Performance Graphics 9-5

 Bar Graphs
Use bar graphs to show changes occurring over time, parts of an entity, relation-
ships between variables, and comparisons.

Use composite-bar graphs to show how parts comprise the entity, and how the
entity relates to other entities.

CPU Monday’s Performance Data

Utilization CPU Utilization vs. Time (Job Type)
90
80
70
60
50
40
30
20
10
0
08:00 10:00 12:00 14:00 16:00 18:00
Time
Press Enter to continue. Batch
F9=Overlay Interact
F3=Exit F6=Print
System
F12=Cancel F16=Save GDF F17=Plot
RV2S074-0

Figure 9-4. Composite-Bar Graph

Floating-bar graphs are similar to composite-bar graphs, except that the first com-
ponent is not shown. Use floating-bar graphs to show the lower limits of each
entity, in addition to the relationship of the elements that comprise the entity.

CPU Monday’s Performance Data

Utilization CPU Utilization vs. Time (Job Type)
90
80
70
60
50
40
30
20
10
0
08:00 10:00 12:00 14:00 16:00 18:00
Time
Press Enter to continue. Batch
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot System
RV2S083-0

Figure 9-5. Floating-Bar Graph

Data Types—Manager Feature

Data types control the number of lines displayed in your graph. They are a means
of categorizing the information provided in your graph. For example, if you want the
graph CPU Over Time and want a separate line plotted for every priority data type,
you would specify *PRIORITY as your data type. You would then be presented
with a display that would allow you to enter 1 to 16 priority ranges for plotting in this
particular graph. Data types, therefore, control the legend entries in your graph.

Data types available for graphing are:

9-6 Performance Tools V4R2

 All Jobs *ALL (default)
Job Type *JOBTYPE
Priority *PRIORITY
Functional Area *FCNARA
IOP (input/output processor) *IOP
Disk *DISK
Communications Lines *CMNLINE

Valid Data Types for Axis Selections

Table 9-2 shows the possible combinations for X-axis and Y-axis values based on
the data type being graphed. For example, if you want to graph Time against Disk
IOP Utilization, specify a data type of *IOP.

Table 9-2 (Page 1 of 3). Valid X-axis and Y-axis Values

X-Axis
Sync Async Total
Total Disk Disk Disk
Trans Nbr I/O Total I/O Total I/O Total
CPU per of Resp per Sync per Async per Disk
Y-Axis Time Util Hour Trans Time Sec I/O Sec I/O Sec I/O
CPU X2 — X1 X1 X1 X1 X1 X1 X1 X1 X1
Util
Trans X2 X1 — X1 X1 X1 X1 X1 X1 X1 X1
per
Hour
Total X2 X1 X1 — X1 X1 X1 X1 X1 X1 X1
Number
of
Trans
Re- X2 X1 X1 X1 — X1 X1 X1 X1 X1 X1
sponse
Time
Sync X2 X1 X1 X1 X1 — X1 X1 X1 X1 X1
Disk
I/O per
Second
Total X2 X1 X1 X1 X1 X1 — X1 X1 X1 X1
Sync
Disk
I/O
Async X2 X1 X1 X1 X1 X1 X1 — X1 X1 X1
Disk
I/O per
Second
Total X2 X1 X1 X1 X1 X1 X1 X1 — X1 X1
Async
Disk
I/O

Chapter 9. Performance Graphics 9-7

 Table 9-2 (Page 2 of 3). Valid X-axis and Y-axis Values
X-Axis
Sync Async Total
Total Disk Disk Disk
Trans Nbr I/O Total I/O Total I/O Total
CPU per of Resp per Sync per Async per Disk
Y-Axis Time Util Hour Trans Time Sec I/O Sec I/O Sec I/O
Total X2 X1 X1 X1 X1 X1 X1 X1 X1 — X1
Disk
I/O per
Second
Total X2 X1 X1 X1 X1 X1 X1 X1 X1 X1 —
Disk
I/O
Com X3 — — — — — — — — — —
IOP Util
Disk X3 — — — — — — — — — —
IOP Util
Local X3 — — — — — — — — — —
Work
Station
IOP Util
Multi- X3 — — — — — — — — — —
function
Com
IOP Util
Multi- X3 — — — — — — — — — —
function
Disk
IOP Util
Disk X4 — — — — — — — — — —
Arm
Util
Disk X4 — — — — — — — — — —
Percent
Occu-
pied
Com X5 — — — — — — — — — —
Line
Util
| Logical X6 X6 X6 X6 X6 X6 X6 X6 X6 X6 X6
| data-
| base
| I/O

9-8 Performance Tools V4R2

 Table 9-2 (Page 3 of 3). Valid X-axis and Y-axis Values
X-Axis
Sync Async Total
Total Disk Disk Disk
Trans Nbr I/O Total I/O Total I/O Total
CPU per of Resp per Sync per Async per Disk
Y-Axis Time Util Hour Trans Time Sec I/O Sec I/O Sec I/O
Key:
1. A graph type of *SCATTER and data type of *ALL, *FCNARA, *JOBTYPE, or *PRIORITY are required.
2. A data type of *ALL, *FCNARA, *JOBTYPE, or *PRIORITY is required.
3. A data type of *IOP is required.
4. A data type of *DISK is required.
5. A data type of *CMNLINE is required.
| 6. A data type of *JOBTYPE and a job type of *DDM are required.

Legends—Manager Feature
The legends displayed in the graph are controlled by the data type specified (for
example, *JOBTYPE). The maximum number of legend entries you can specify for
each data type is as follows:
Data Type Maximum Legend Entries
All 1
Job Type 16
Priority 16
Functional Area 16
IOP 2
Disk 2
Communications Line 16

Create Graph Package—Manager Feature

To create a new graph package, type a 6 (Create graph package), the graph
package name, and the text description on the first line under the Option,
Format/Package, and Text columns, and press the Enter key. The Create Graph
Package display appears.

Chapter 9. Performance Graphics 9-9

 à ð
Create Graph Package

Graph package . . : PACKAGE2

Library . . . . : QPFRDATA

Type options, press Enter.

1=Select 5=Display sample graph

Option Format Text

_ FORMAT1 CPU Utilization vs. Time-Functional Areas
_ FORMAT2 Response Time vs. Time-Functional Areas
_ QIBMASYNC Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP Communications IOP Utilization vs. Time
_ QIBMCPUPTY CPU Utilization vs. Time (Priority)
_ QIBMCPUTYP CPU Utilization vs. Time (Job Type)
_ QIBMDSKARM Disk Arm Utilization vs. Time
_ QIBMDSKIOP Disk IOP Utilization vs. Time
_ QIBMDSKOCC Percentage of Disk Occupied vs. Time
_ QIBMLWSIOP Local Workstation IOP Utilization vs. Time
_ QIBMMFCIOP Multifunction IOP (Comm) Util vs. Time
More...

F3=Exit F5=Refresh F12=Cancel F16=Sort by text

á ñ
On this display, type a 1 (Select) by any graph formats that you want to include in
the graph package. If you are unsure about including a graph format in the
package, type a 5 (Display sample graph) by the format in question. This displays a
sample graph using the format selected. When you have made all of your
selections and there are only 1’s in the Option column, press the Enter key to
create the graph package.

Change Graph Formats and Packages—Manager Feature

To change an existing graph format or graph package, type a 2 (Change) next to
the format or package name on the Work with Graph Formats and Packages
display, and press the Enter key. If you are changing a graph format, the Change
Graph Format (CHGGPHFMT) command prompt appears. Make your changes and
press the Enter key. If you are changing a graph package, the Change Graph
Package display appears.

9-10 Performance Tools V4R2

 à ð
Change Graph Package

Graph package . . : PACKAGE1

Library . . . . : QPFRDATA
Text . . . . . : Text for package 1

Type options, press Enter.

1=Select 5=Display sample graph

Option Format Text

1 FORMAT1 CPU Utilization vs. Time-Functional Areas
1 FORMAT2 Response Time vs. Time-Functional Areas
1 QIBMASYNC Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP Communications IOP Utilization vs. Time
_ QIBMCPUPTY CPU Utilization vs. Time (Priority)
_ QIBMCPUTYP CPU Utilization vs. Time (Job Type)
_ QIBMDSKARM Disk Arm Utilization vs. Time
_ QIBMDSKIOP Disk IOP Utilization vs. Time
_ QIBMDSKOCC Percentage of Disk Occupied vs. Time
_ QIBMLWSIOP Local Workstation IOP Utilization vs. Time
More...
F3=Exit F5=Refresh F1ð=Restore list F12=Cancel
F15=Sort by format F16=Sort by text

á ñ
On this display, 1’s appear next to the graph formats that are already included in
the graph package. To eliminate a graph format from the package, replace the 1
with a blank. To add additional graph formats to the package, type a 1 (Select) next
to the graph formats you want to include. To display a sample of a graph format,
type a 5 (Display sample graph) next to the graph format and press the Enter key.
A sample graph using the graph format is displayed.
Note: You cannot change the IBM standard graph formats and graph package
(QIBMxxxxxx).

Copy Graph Formats and Packages—Manager Feature

To copy a graph format or graph package, type a 3 (Copy) next to the format or
package name on the Work with Graph Formats and Packages display and press
the Enter key.

à ð
Work with Graph Formats and Packages

Library . . . . . . . . . . . QPFRDATA

Type options, press Enter.

1=Create graph format 2=Change 3=Copy 4=Delete
5=Display sample graph 6=Create package 8=Display package contents

Either the Copy Graph Format (CPYGPHFMT) or Copy Graph Package

(CPYGPHPKG) command prompt appears. You can copy a graph format or
package to another library or into the same library under a different name. A graph
format or package that is created in a library cannot have the same name as a
graph format or graph package that already exists in the library.

Copying graph formats and packages is useful for changing a base format or
package, such as the IBM standard graph formats and package (QIBMxxxxxx).

Chapter 9. Performance Graphics 9-11

Delete Graph Formats and Packages—Manager Feature
To delete graph formats and graph packages, type a 4 (Delete) next to the format
and package names on the Work with Graph Formats and Packages display, and
press the Enter key.

If a graph format you selected to delete is contained in any graph packages, a

warning message displays telling you that the format is in a package. If you delete
the graph format, the format is also removed from the graph package. If all the
graph formats in a graph package are deleted, the package is also deleted.
Note: You cannot delete the IBM standard graph formats and graph package
(QIBMxxxxxx).

Display Sample Graph

To display a sample of a graph format, type a 5 (Display sample graph) next to the
graph format name on the Work with Graph Formats and Packages display, and
press the Enter key. A sample graph using the graph format is displayed.
Note: This option is not valid for graph packages.

Display Package Contents—Manager Feature

To display the contents of a graph package, type an 8 (Display package contents)
next to the graph package name on the Work with Graph Formats and Packages
display, and press the Enter key. The Display Package Contents display appears.
Note: Option 8 (Display package contents) cannot be specified for graph formats.

à Display Package Contents

ð
Graph package . . : PACKAGE1
Library . . . . : QPFRDATA

Type options, press Enter.

5=Display sample graph

Option Format Text

_ FORMAT1 CPU Utilization vs. Time-Functional Areas
_ QIBMASYNC Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP Communications IOP Utilization vs. Time

Bottom

F3=Exit F5=Refresh F12=Cancel F16=Sort by text

á ñ
On this display, type a 5 (Display sample graph) to see a sample graph displayed
using the graph format.

9-12 Performance Tools V4R2

 Work with Historical Data—Manager Feature
The Display Historical Graph (DSPHSTGPH) command uses historical data to
show the changes in resource utilization on your system over time. Historical data
is a summary of the performance data created by the performance monitor.

Notes:
1. Files are created to contain the historical data. For each performance member
with historical data, there is a single value for each type of information that can
be graphed for each day of the member’s performance collection period. Thus,
the amount of data is reduced and summarized into the historical files. Once
you have historical data for a member, you may choose to delete the perform-
ance data (DLTPFRDTA) created through the initial performance data collection
to free file storage space.
| 2. If you want to display a historical graph, select a performance data member
| that contains less than 400 intervals.
| 3. Any time a collection extends beyond midnight, each day counts as one
| member.

Because historical graphs can help show trends in your system’s performance, it is
recommended that you create historical data in a given library for members that are
collected at the same time. (For example, you might want to compare data that was
all collected on Wednesdays from 8:00 a.m. to 12:00 p.m., whereas you probably
would not want a historical graph with one member collected on Wednesday from
8:00 a.m. to 12:00 p.m. and the other on Saturday from 1:00 to 5:00 p.m.)

If you select option 2 (Work with historical data) on the Performance Tools
Graphics menu, the Work with Historical Data display appears.

à ð
Work with Historical Data

Library . . . . . . QPFRDATA

Type options, press Enter.

1=Create historical data 4=Delete historical data

Member Historical
Option Name Data Date Time
_ Q95318ð843 NO 11/14/95 ð8:43:15
_ Q953171ð5ð NO 11/13/95 1ð:51:ðð
_ SATDATA YES 11/11/95 1ð:42:48
_ TESTDATA YES 11/11/95 1ð:26:12
_ NOV111995 NO 11/11/95 ð9:57:27
_ Q95315ð955 NO 11/11/95 ð9:55:41
_ FRIDAY YES 11/1ð/95 11:17:ð3
_ Q953132332 YES 11/ð9/95 23:32:19
_ Q9531214ð7 YES 11/ð8/95 14:ð7:11
_ Q953121142 NO 11/ð8/95 11:42:3ð
_ Q953111538 NO 11/ð7/95 15:39:ð2
More...
F3=Exit F5=Refresh F11=Display text F12=Cancel
F15=Sort by member F16=Sort by text

á ñ
The member name, a historical data indicator, and the date and time you collected
each set of performance data appear on this display. To display the member text

Chapter 9. Performance Graphics 9-13

 description, press F11 (Display text). If you cannot find the data you want to work
with, use the appropriate function key to sort the sets of performance and historical
data. You can sort them by member name, text description, or by the date and time
the member was created. When you find the data you want to work with, indicate
the function you want to perform by typing the appropriate option.

If you are searching for performance or historical data located in a library that is
different from the one currently listed in the Library field at the top of the display,
type a new library name in the Library field and press the Enter key. A list of per-
formance and historical data members available in the library you specified
appears. You can then select one of them to work with.
Note: All of the members in the historical data must have unique names. If you
create a member that has the same name as a historical data member, you
may want to change the name by using the Copy Performance Data
(CPYPFRDTA) command to use the new member for historical purposes.

It is best to use the created name option (*GEN) on the STRPFRMON command to
make sure that the names of your performance data members are unique.

Create Historical Data

To create historical data for performance members, type a 1 (Create) by the
members, and press the Enter key. The Confirm Create of Historical Data display
appears.

à ð
Confirm Create of Historical Data

Library . . . . . : QPFRDATA

Press Enter to confirm your choices for 1=Create.

Press F12=Cancel to return to change your choices.

Member Historical
Option Name data Date Time
1 Q95318ð843 NO 11/14/95 ð8:43:15
1 Q953171ð5ð NO 11/13/95 1ð:51:ðð
1 SATDATA YES 11/11/95 1ð:42:48

Bottom
F11=Display text F12=Cancel

á ñ
On this display, press the Enter key to create historical data for the members. Once
historical data has been created for a member, you can delete the original perform-
ance data using the Delete Performance Data (DLTPFRDTA) command if the data
is not needed for performance analysis, capacity planning, or performance
graphing.

9-14 Performance Tools V4R2

Delete Historical Data
To delete the historical data created by the Create Historical Data command, type a
4 (Delete) by members that contain historical data, and press the Enter key. This
does not delete the original performance data.
Note: If the performance data for a member no longer exists, you cannot re-create
historical data for that member after the historical data has been deleted.

Display Graphs and Packages—Manager Feature

You can view, print, or plot graphs from your display. You can also store a graph in
a GDF file for use by other utilities, such as the BGU. This is done on the Specify
Graph Options display.

If you select option 3 (Display Graphs and Packages) on the Performance Tools
Graphics menu, the Display Graphs and Packages display appears.

à ð
Display Graphs and Packages

Select one of the following:

1. Display performance data graphs

2. Display historical data graphs

Selection or command
===> __________________________________________________________________
_______________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ
Two distinct types of graphs can be displayed: performance graphs and historical
graphs. Performance graphs use performance data collected from a single run of
the performance monitor. Performance graphs are used to single out jobs that are
performing poorly or to evaluate which activities were performed by a user or class
of users on the system during a specified period.

Historical graphs use performance data collected from several runs of the perform-
ance monitor. Historical data is the summary of the performance data created by
the STRPFRMON command. The CRTHSTDTA command is used to summarize
the performance data for use by the historical graphs. Historical graphs are used to
show how the performance of a system has changed over time.
Note: It is best to collect the performance data used for historical graphs over the
same period of time. For example, if your normal working day is from 8:00
a.m. to 5:00 p.m., you would not want to create a historical graph to eval-

Chapter 9. Performance Graphics 9-15

 uate system performance during working hours using system performance
data collected from 5:00 p.m. to 8:00 a.m.

If you want to collect data at predetermined times, use the Work with Performance
Collection (WRKPFRCOL) command to schedule the running of the performance
monitor. See Chapter 3, Collecting System Performance Data, for more information
on this command.

Display Performance Graphs—Manager Feature

If you select option 1 (Display performance data graphs) on the Display Graphs and
Packages display, the Select Graph Formats and Packages display appears.

à ð
Select Graph Formats and Packages

Library . . . . . . . . . . . QPFRDATA

Type options, press Enter.

1=Select 5=Display sample graph 8=Display package contents

Format/
Option Package Type Text
_ NEWPACKAGE PACKAGE Graph Package for Job Types
_ PACKAGE1 PACKAGE Graph Package containing IOP formats
_ QIBMPKG PACKAGE IBM Graph Package
_ FORMAT1 FORMAT CPU Utilization vs. Time-Functional Areas
_ FORMAT2 FORMAT Response Time vs. Time-Functional Areas
_ QIBMASYNC FORMAT Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP FORMAT Communications IOP Utilization vs. Time
_ QIBMCPUPTY FORMAT CPU Utilization vs. Time (Priority)
_ QIBMCPUTYP FORMAT CPU Utilization vs. Time (Job Type)
_ QIBMDSKARM FORMAT Disk Arm Utilization vs. Time

More...
F3=Exit F5=Refresh F12=Cancel F14=Sort by format F15=Sort by text

á ñ
This display shows you the graph formats and graph packages that exist in the
library you specified. The graph format or graph package name, a format or
package indicator, and a text description appear on the display. If you cannot find
the format or package you want to use in your performance graph, use the appro-
priate function key to sort the formats and packages. You can sort them by name,
type, or text description. When you find the graph format or package you want to
use in your performance graph, type a 1 in the corresponding Option field.

If you are searching for a graph format or graph package located in a library that is
different from the one currently listed in the Library field at the top of the display,
type a new library name in the Library field, and press the Enter key. A list of graph
formats and graph packages available in the library you specified appears. You can
then select one of them to use in your performance graph.
| Note: If you want to display a historical graph, select a performance data member
| that contains less than 400 intervals.

9-16 Performance Tools V4R2

Display Sample Graph—Manager Feature
To display a sample of a graph format, type a 5 (Display sample graph) next to the
graph format, and press the Enter key. A sample graph using the graph format
appears.
Note: This option is not valid for graph packages.

Display Graph Package—Manager Feature

To display the contents of a graph package, type an 8 (Display package contents)
next to the graph package, and press the Enter key. A list of the graph formats
contained in the graph package appears.
Note: This option is not valid for graph formats.

Select Performance Data Member—Manager Feature

After you select a graph format or graph package to use in your performance
graph, the Select Performance Data Member display appears.

à ð
Select Performance Data Member

Library . . . . . . QPFRDATA

Type options, press Enter.

1=Select

Member
Option Name Text Date Time
_ Q95318ð843 11/14/95 ð8:43:15
_ Q953171ð5ð 11/13/95 1ð:51:ðð
_ SATDATA3 Saturday Data-third run 11/11/95 1ð:42:48
_ SATDATA2 Saturday Data-second run 11/11/95 1ð:26:12
_ SATDATA1 Saturday Data-first run 11/11/95 ð9:57:27
_ Q95315ð955 11/11/95 ð9:55:41

More...
F3=Exit F5=Refresh F12=Cancel F15=Sort by member
F16=Sort by text

á ñ
The member name, a text description, and the date and time you collected each
set of performance data appear on this display. If you cannot find the data you
want to display, use the appropriate function keys to sort the sets of performance
data. You can sort the data by member name, text description, or by the date and
time the member was created. When you find the performance data you want to
use in your performance graph, type a 1 in the corresponding Option field.

If you are searching for a member located in a library that is different from the one
currently listed in the Library field at the top of the display, type a new library name
in the Library field and press the Enter key. A list of the performance members
available in the specified library appears. You can then select a member to display.

Chapter 9. Performance Graphics 9-17

Select Categories for Performance Graphs—Manager Feature
If the graph format or graph package you previously selected does not graph only
IOP, disk, or communications line data, the Select Categories for Performance
Graphs display appears.

à ð
Select Categories for Performance Graphs

Member . . . . . : MONDAYDATA
Library . . . . : QPFRDATA

Type options, press Enter. Press F6 to include all data in the graph.
1=Select

Option Category
_ Job
_ User ID
_ Subsystem
_ Pool
_ Communications line
_ Control unit
_ Functional area

Bottom
F3=Exit F6=Include all data in the graph F12=Cancel

á ñ
Type a 1 in the Option column next to the categories of information from which you
want performance data. Press the Enter key.
Note: Normally, you include all categories of information in your graph. To do this,
do not type a 1 in any category. Instead, simply press F6.

If you choose to display the graph with only certain categories of information, a
display appears that allows you to enter selection criteria for each category. This is
the same as selecting categories of information to include in performance reports.
See Chapter 7, Performance Reports—Manager Feature, for more information on
selecting and omitting.

Specify Graph Options—Manager Feature

When you have chosen the information you want to appear on your performance
graph, or if you selected a graph format with IOP, disk, or communications line data
type, the Specify Graph Options display appears.

9-18 Performance Tools V4R2

à ð
Specify Graph Options

Type choices, press Enter.

Graph title . . . \MBRTEXT

Graph subtitle . . CPU Utilization vs. Time

X-axis range:
First . . . . . \AUTO___ \SAME, \AUTO, Number
Last . . . . . . ________ Number

Y-axis range:
First . . . . . \AUTO___ \SAME, \AUTO, Number
Last . . . . . . ________ Number

Area fill . . . . \NO \SAME, \YES, \NO

Start:
Day . . . . . . \FIRST \FIRST, MM/DD/YY
Time . . . . . . \FIRST \FIRST, HH:MM:SS
More...
F3=Exit F12=Cancel

á ñ
Figure 9-6. Specify Graph Options

Page down to view the rest of the graph options.

à ð
Specify Graph Options

Type choices, press Enter.

Stop:
Day . . . . . . \LAST \LAST, MM/DD/YY
Time . . . . . . \LAST \LAST, HH:MM:SS

Output . . . . . . \_______ \, \PRINT, \PLOT, \OUTFILE

Bottom
F3=Exit F12=Cancel

á ñ
On this display you can specify a new title, subtitle, axis ranges, area fill value, start
time and date, stop time and date, and output value for your performance graph. If
you selected a graph format for your performance graph, the values for the title,
subtitle, axis ranges, and area fill defined in the graph format appear. Changing any
of the values on the Specify Graph Options display only changes the format for the
graph created. The graph format does not change. If you selected a graph package
for your performance graph, *SAME appears for the title, subtitle, and axis ranges.

Chapter 9. Performance Graphics 9-19

 *SAME means to leave these values as they are defined in the individual graph
formats in the package. If you specify any new values, the new values appear on
all of the graphs in the package.

For example, if you type New Graph Title for the graph title and the graph package
contained three graph formats, the resulting three graphs would have “New Graph
Title” as their title.

The area fill option allows you to override the area fill option on the graph format to
display a graph more quickly. Filling in (or shading) an area is accomplished by
drawing several lines. Densely shaded patterns require more lines. Each line that is
drawn takes time. Consequently, the graph displays faster if area fill is not used. If
the area fill option on the graph format is *YES, then selecting *NO for the area fill
option causes the area not to be filled.

You may specify the start and stop date and time for the performance data to be
shown in the graph. If you do not specify the start and stop date and time, the
graph includes data from the first (or only) date that data was collected to the last
(or only) date that data was collected.

The output option specifies how the graphs are to be displayed.

Press the Enter key to display your graph or graphs.

Display Historical Graphs—Manager Feature

Historical graphs allow you to graphically see how your system performed during
many runs of the performance monitor. This shows you how the performance of
your system has changed over time. For example, it can show how processing unit
utilization increased or fluctuated.

If you select option 2 (Display historical data graphs) on the Display Graphs and
Packages display, the Select Graph Formats and Packages display appears. This
is the same display that is shown for displaying performance graphs. (See “Display
Performance Graphs—Manager Feature” on page 9-16 for more information.) After
you select a graph format or graph package from the Select Graph Formats and
Packages display, the Specify Graph Options display appears.

9-20 Performance Tools V4R2

 à ð
Specify Graph Options

Type choices, press Enter.

Graph title . . . \BLANK___________________________________________

Graph subtitle . . \BLANK___________________________________________

X-axis range:
First . . . . . \AUTO___ \SAME, \AUTO, Number
Last . . . . . . ________ Number

Y-axis range:
First . . . . . \AUTO___ \SAME, \AUTO, Number
Last . . . . . . ________ Number

Area fill . . . . \NO \SAME, \YES, \NO

Data library . . . QPFRDATA__ Name

More...
F3=Exit F12=Cancel

á ñ
Page down to view the rest of the graph options.

à ð
Specify Graph Options

Type choices, press Enter.

Start:
Day . . . . . . . . \FIRST__ \FIRST, \SELECT, MM/DD/YY

Stop:
Day . . . . . . . . \LAST___ \LAST, MM/DD/YY

Create historical
data . . . . . . . . . \NO_ \YES, \NO

Output . . . . . . . . \_______ \, \PRINT, \PLOT, \OUTFILE

Bottom
F3=Exit F12=Cancel

á ñ

Display Graph Overlay—Manager Feature

Once you have a performance graph or historical graph displayed, you can define
one overlay by pressing F9 (Overlay). An overlay is a graph that is placed on top of
another graph so that you can see both graphs at the same time. Overlays can
help you compare one graph to another as shown below.

You must select a graph format with the same X-axis specified.

Chapter 9. Performance Graphics 9-21

 If you want to overlay a historical graph, you cannot display a graph format with
functional area data type.

Note that when you overlay a graph, there is a maximum of 16 legend entries
between the two graphs. Therefore, if you are currently displaying a graph with two
legend entries, your overlaid graph may have only a maximum of 14 legend entries
(if allowed for the data type in the graph format). See “Legends—Manager Feature”
on page 9-9 for the maximum number of legend entries for the individual data
types. If you are currently displaying a graph with 16 legend entries, you cannot
overlay a second graph.

Press F9 (Overlay), and the Select Graph Format display appears. Select the graph
format that you want to overlay above the graph that is currently displayed.

à ð
Select Graph Format

Library . . . . . . QPFRDATA

Type option, press Enter.

1=Select

Option Format Text

_ QIBMASYNC Asynchronous Disk I/O per Second vs. Time
_ QIBMCMNIOP Communications IOP Utilization vs. Time
_ QIBMCPUPTY CPU Utilization vs. Time (Priority)
_ QIBMCPUTYP CPU Utilization vs. Time (Job Type)
_ QIBMDSKARM Disk Arm Utilization vs. Time
_ QIBMDSKIOP Disk IOP Utilization vs. Time
_ QIBMDSKOCC Percentage of Disk Occupied vs. Time
_ QIBMLWSIOP Local Workstation IOP Utilization vs. Time
_ QIBMMFCIOP Multifunction IOP (Comm) Util vs. Time
_ QIBMMFDIOP Multifunction IOP (Disk) Util vs. Time
_ QIBMRSP Interactive Response Time vs. Time
More...

F3=Exit F5=Refresh F12=Cancel F16=Sort by text

á ñ
Select a graph format and press the Enter key, and the Specify Graph Overlay
Options display appears.

9-22 Performance Tools V4R2

à ð
Specify Graph Overlay Options

Type choices, press Enter.

New graph title \BLANK________________________________________

New graph subtitle \BLANK________________________________________

Y-axis range:
First . . . . . \AUTO_____ \SAME, \AUTO, Number
Last . . . . . . __________ Number

Area fill . . . . \NO_ \YES, \NO

F3=Exit F12=Cancel

á ñ
Specify a title and subtitle for your new, overlaid graph on this display. If you do
not specify a new title and subtitle, your new graph title and subtitle are left blank.

The Y-axis range value defaults to the value that was specified in the graph format.
Here, again, you have the chance to change it. You may choose to have the same
range as defined in the graph format (*SAME), you may choose to have it automat-
ically fit the range of values (*AUTO), or you may specify the range yourself by
typing in the numbers.

You also select whether to have area fill in the overlaid graph.

After you press the Enter key, your two graphs should be displayed. You can use
the function keys on the display to print or plot the overlay or send the overlay
format to a GDF file. Figure 9-7 shows an example of an overlay graph.

CPU Monday’s Performance Data CPU Transactions

Utilization Utilization and Trans per per Hour
90 Hour vs. Time 3200
80 2800
70 2400
60 2000
50 1600
40 1200
30 800
20 400
10 0
0

08:00 10:00 12:00 14:00 16:00 18:00

Time
Press Enter to continue. Batch
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot System

RV2S077-0

Figure 9-7. Overlay Graph Example

Chapter 9. Performance Graphics 9-23

9-24 Performance Tools V4R2
Chapter 10. Performance Utilities—Manager Feature
This chapter describes the commands that you access with option 5 (Performance
utilities) on the IBM Performance Tools menu. When you choose option 5, the Per-
formance Utilities display appears.

à ð
Performance Utilities
Select one of the following:

1. Work with job traces

2. Work with Performance Explorer
3. Select file and access group utilities

The utilities shown on the Performance Utilities display provide you with support for
the detailed performance analysis of applications when you are working to under-
stand or improve the performance of those applications.

See “Summary of Data Collection and Report Commands—Manager Feature” on

page 3-13 for an overview of the commands you use with Performance Tools, their
data collection requirements, and their intended uses. For a description of the per-
formance explorer function (option 2) see Chapter 11, Performance Explorer.

Job Traces
If you select option 1 (Work with job traces) on the Performance Utilities display,
the Work with Job Traces display appears.

à ð
Work with Job Traces

Select one of the following:

1. Start job trace

2. Stop job trace
3. Print job trace reports

On this display you can choose to start or stop a job trace. After you collect the
trace data, you can print job trace reports that show information about input/output
(I/O) operations, file use, transaction timing, job flow, and so on.

The options in the Job Trace display and the corresponding commands are as
follows:
Job Traces Corresponding Command
Start Job Trace STRJOBTRC
Stop Job Trace ENDJOBTRC
Print Job Trace Reports PRTJOBTRC

 Copyright IBM Corp. 1998 10-1

 For more information on job traces, see “Analyzing Job Flow and Transaction
Performance” on page 10-2.

File and Process Access Group (PAG) Utilities

If you choose option 3 (Select file and access group utilities) on the Performance
Utilities menu, the Select File and Access Group Utilities display appears.

à ð
Select File and Access Group Utilities

Select one of the following:

1. Analyze program/file use

2. Analyze physical/logical file relationships
3. Analyze file key structure
4. Collect/display access group data
5. Analyze access group data

On this display you can choose to create reports that show the program-to-file use,
the physical-to-logical file relationships, the file key structure, or access group data.
You can also use this display to determine if the application programs use shared
display and database files, if the files are ordered by their frequency of use, if the
files remain open but have no activity, or if programs free their static storage or
keep it active.

Notes:
1. Before you use option 3, be sure that the processing for option 2 has com-
pleted. The output from option 2 is used as input for this function.
2. Option 5 is dependent on data collected by option 4. So you must take option 4
first.

The options in the Select File and Access Group Utilities display and the corre-
sponding commands are as follows:

File and Access Group Utilities Corresponding

Command
Analyze Program/File Use ANZPGM
Analyze Physical/Logical File Relationships ANZDBF
Analyze File Key Structure ANZDBFKEY
Collect/Display Access Group Data DSPACCGRP
Analyze Access Group Data ANZACCGRP

For more information on file and PAG utilities, see “Analyzing the Relationship of
Programs and Database Files” on page 10-9 and “Analyzing Process Information”
on page 10-16.

Analyzing Job Flow and Transaction Performance

Use the job trace commands to collect trace information about a job. You can do
this while the job runs in the normal production environment, or you can set up a
special test for a job or program and trace how it runs. Once you collect the trace
information, print the reports (there are two summary reports and one detail report).
The summary reports allow you to determine the overall performance of the job

10-2 Performance Tools V4R2

 without analyzing the detail report. Use the summary reports to guide you through
the detail report.

Do not produce the detailed job analysis until you define which program or job you
want to analyze.

Job trace analysis enhances the operating system’s standard trace job reports and
provides a summary of job operation and transaction processing. The primary use
for job trace analysis is to determine how a job processes. You can determine
what parts of a job use the most resources, and measure the effect of program
changes relative to previous trace data. Do not use job trace analysis to determine
accurate job or transaction processing times.

Start Job Trace (STRJOBTRC) Command

Use the STRJOBTRC command to start the job trace function. The End Job Trace
(ENDJOBTRC) and Print Job Trace (PRTJOBTRC) commands provide summary
and detail reports of the job trace data.

Consider the following points when you use STRJOBTRC:

Ÿ The job trace function usually changes the paging characteristics of a job.
Therefore, the trace reports may not show representative times for program
operation.
Ÿ To cancel the job trace without saving any of the collected data, use the
TRCJOB SET(*END) command.
Ÿ The job trace function issues a Start Service Job (STRSRVJOB) command if a
job other than the current job is specified on the STRJOBTRC command.

End Job Trace (ENDJOBTRC) Command

Use the ENDJOBTRC command to do the following:
Ÿ Stop the job trace and direct the trace data into a user-defined database file
member.
Ÿ You may start the PRTJOBTRC command to print the reports that analyze the
trace data. These analysis reports provide an estimate of the response and
processing times. They also show the number of database reads, nondatabase
reads, and write I/O operations.

The database file QAPTTRCJ is created as output when you use the ENDJOBTRC
command. Table 10-1 shows the names and descriptions of the fields in the
QAPTTRCJ file.

Table 10-1 (Page 1 of 2). QAPTTRCJ File

Field Name Description
SCFUNC Type of function
SCSTYP Subtype of function
SCFLD1 Column heading 1
SCFLD2 Column heading 2
SCFLD3 Column heading 3
SCTIME Time of trace record

Chapter 10. Performance Utilities—Manager Feature 10-3

 Table 10-1 (Page 2 of 2). QAPTTRCJ File
Field Name Description
SCSEQ Record sequence number
SCENT Entry machine interface (MI) instruction number
SCEXT Exit machine interface (MI) instruction number
SCINV Call level
SCCPU CPU time used
SCDB Database reads
SCPAG Nondatabase reads
SCWAIT Number of waits
SCRCEN Century. 0 indicates the twentieth century. 1 indicates the twenty-
first century.
SCRDAT Date
SCRTIM Time
SCSYNM System name

The printer file created by this command is the same as that created by the
PRTJOBTRC command, as described in Print Job Trace (PRTJOBTRC) Command.

Print Job Trace (PRTJOBTRC) Command

Use the PRTJOBTRC command to print a report of all, or a selected part, of the job
trace data. The job trace data that prints comes from the database file member that
was created when you ran the ENDJOBTRC command.
Note: There may be gaps in the sequence numbers on the report. These are
caused by undefined trace records that may contain unprintable characters.
To view these records, use the TRCJOB report, which gives you a
hexadecimal display of these fields.

The following printer files are the output when you use the PTRJOBTRC command:
File Description
QPPTTRC1 First part of the summary report (Trace Analysis Summary)
QPPTTRC2 Second part of the summary report (Trace Analysis I/O Summary)
QPPTTRCD Trace record detail report (Trace Job Information)

Both the Trace Analysis Summary Report and the Trace Analysis I/O Summary
Report show the job trace data detail by transaction. On these reports, two lines for
each transaction show all the trace records for that transaction. A transaction
boundary is determined by consecutive trace records with these characteristics:
Ÿ The first trace record indicates a call to the program specified by the end of
transaction (ENDTNS) parameter.
Ÿ The second trace record indicates a return to the program specified by the start
of transaction (STRTNS) parameter.

The default ENDTNS and STRTNS parameters cause the trace records to be
shown by work station transactions on these reports. A transaction begins when a
user presses the Enter key, or otherwise responds to a program prompt, and ends

10-4 Performance Tools V4R2

when the program next requests input from the work station. You can change these
parameters in order to summarize other types of transactions, such as record proc-
essing (useful when tracing a batch job), or communications I/O.

The summary reports show you the number and types of I/O operations that
occurred for each transaction, the number of full and shared file opens and closes,
the number of subfile operations, and the number of messages that occurred in the
transaction. Messages may be the result of normal operation or they may be due to
program actions that you can avoid (full open/close, duplicate keys in a file, or
incorrect subfile processing).

The summary reports also contain a reference to the detail report. Every detail
record has a sequence number in it. The summaries show the starting and ending
detail report sequence numbers for each transaction summarized. The detail report
program can be limited to a range of sequence numbers. This feature allows you to
run the summaries, then print only the detail you are interested in.

The collection of trace data takes a certain amount of processing time, the amount
of which can vary depending on such factors as system load and model. This
overhead time is included in the trace data on which the PRTJOBTRC command
reports. The command attempts to subtract the overhead time from the reported
figures, leaving only the time used for program processing. Due to the variability of
the overhead time, this adjustment may not be accurate. This adjustment is an
estimate only. Therefore, do not use reported processing times as an absolute
measure of the response time of a program or set of programs.

Figure 10-1 on page 10-6 shows an example of the Trace Analysis Summary
Report.

The header of the Trace Analysis Summary Report shows the following values:
Title The title specified on the command.
FILE The name of the database file containing the trace data.
LIBRARY The library the database file is in.
MBR The database file member containing the trace data.
JOB The name of the job that was traced.

The columns in the detailed section of the Trace Analysis Summary Report are as
follows:
ACTIVE or WAIT-ACT The time between the ENDTNS and STRTNS programs is
labeled Wait-Act. If you were tracing an interactive job and used the
default STRTNS and ENDTNS parameters, this value is the time taken
to process the transaction.
SECONDS The approximate time the job was waiting or active.
CPU SECONDS The approximate processing unit time used for the transaction. If
the value is zero (or blank), you may have chosen the wrong value for
the model parameter.
DB READS The number of physical database reads that occurred.
NON-DB RDS The number of physical nondatabase reads that occurred.
WRITES The number of physical writes that occurred.

Chapter 10. Performance Utilities—Manager Feature 10-5

 Title TRACE ANALYSIS SUMMARY 12/ð1/95
FILE-QAJOBTRC LIBRARY-QPFRDATA MBR-QAJOBTRC JOB- BYSINN .VLLXR239 .ðð3368
P H Y S I C A L I / O
SECONDS CPU SECONDS DB READS NON-DB RDS WRITES WAITS SEQUENCE
WAIT-ACT 4.852 .ðð9 16
ACTIVE 1.425 .788 33 5 1 1ð8
WAIT-ACT 4.ð93 .ð17 3 112
ACTIVE .247 .11ð 7 5 1 119
WAIT-ACT 3.736 .ðð9 123
ACTIVE .658 .572 8 5 1 18ð
WAIT-ACT 1.793 .ðð5 184
ACTIVE .512 .193 19 3 2 2ð6
WAIT-ACT 4.195 .ðð9 21ð
AVERAGE .711 .426 18 5 1 4
TOTAL 2.842 1.7ð3 7ð 18 5

Figure 10-1. Trace Analysis Summary Report

WAITS The number of waits that occurred.

SEQUENCE The job trace sequence number in the detail report that this summary
line refers to.
AVERAGE and TOTAL Averages and totals for the fields described above. The
entry on the Average line in the Sequence column shows the number of
STRTNS and ENDTNS pairs encountered. For an interactive job, this is
the number of transactions entered while the trace was on if the default
STRTNS and ENDTNS values were used.

Figure 10-2 on page 10-7 shows an example of the Trace Analysis I/O Summary
Report.

The columns in the Trace Analysis I/O Summary Report are as follows:
Title The title specified on the command.
FILE The name of the database file containing the trace data.
LIBRARY The library the database file is in.
MBR The database file member containing the trace data.
JOB The name of the job that was traced.
WAIT-ACT The time that the job was inactive, probably due to typing or think time
by the user.
ACTIVE The time the job was processing.
SECONDS The time the job was waiting or active.
SEQNCE The job trace sequence number in the detail computer printout that this
summary line refers to.
PROGRAM NAME The name of the last program called that was not in the library
QSYS before the end of a transaction.
PROGRAM CALL The number of non-QSYS library programs called during the
step. This is not the number of times that the program named in the
PROGRAM NAME field was called.
PROGRAM INIT The number of times that the IBM-supplied initialization program
was called during the transaction. For RPG programs this is QRGXINIT,
for COBOL it is QCRMAIN. Each time the user program ends with LR
(RPG) or END (COBOL), the IBM-supplied program is also called. This
is not the number of times the program named in the Program Name

10-6 Performance Tools V4R2

 Title TRACE ANALYSIS I/O SUMMARY 12/ð1/95
FILE-QAJOBTRC LIBRARY-QPFRDATA MBR-QAJOBTRC JOB- BYSINN .VLLXR239 .ðð3368
P R O G R A M \\\\\\\ PROGRAM DATA BASE I/O \\\\\\\ FULL SHARE SUBFILE
SECONDS SEQNCE NAME CALL INIT GETDR GETSQ GETKY GETM PUT PUTM UDR OPN CLS OPN CLS READS WRITES MSGS
WAIT-ACT 4.852 16
ACTIVE 1.425 1ð8 QPTPAGDð 1 1 11
WAIT-ACT 4.ð93 112
ACTIVE .247 119
WAIT-ACT 3.736 123
ACTIVE .658 18ð 11
WAIT-ACT 1.793 184
ACTIVE .512 2ð6 1
WAIT-ACT 4.195 21ð
AVERAGE .711 4 6
TOTAL 2.842 1 1 1 22

Figure 10-2. Trace Analysis I/O Summary Report

field was initialized. QCRMAIN is used for functions other than program
initialization (for example, blocked record I/O, some data conversions).
PROGRAM DATABASE I/O The number of times the IBM-supplied database
modules were used during the transaction. The database module names
have had the QDB prefix removed (PUT instead of QDBPUT). The type
of logical I/O operation performed by each is as follows:
GETDR Get direct
GETSQ Get sequential
GETKY Get by key
GETM Get multiple
PUT, PUTM Add a record
UDR Update, delete, or release a record

The values for OPENS and CLOSES in the programs are as follows:
FULL OPN The number of full opens for all types of files.
FULL CLS The number of full closes for all types of files.
SHARE OPN The number of shared opens for all types of files.
SHARE CLS The number of shared closes for all types of files.

The valid values for Subfile I/O are as follows:

SUBFILE READS The number of subfile reads.
SUBFILE WRITES The number of subfile writes.
MSGS The number of messages sent to the job during each transaction.

The Trace Job Information Report, shown in Figure 10-3 on page 10-8, has essen-
tially the same format as the system-supplied trace job output. The AS/400
Licensed Internal Code Diagnostic Aids - Volume 1 book contains additional infor-
mation on trace jobs.

The columns in the Trace Job Information Report are as follows:

TIME The time of day for the trace entry. The time is sequentially given in
hours, minutes, seconds, and fractions of a second.
SEQNBR The number of the trace entry.

Chapter 10. Performance Utilities—Manager Feature 10-7

 Sample Job Trace Report TRACE JOB INFORMATION PAGE 1
FILE-QAJOBTRC LIBRARY-QPFRDATA MBR-QAJOBTRC JOB- BYSINN .VLLXR239 .ðð3368
TIME SEQNBR FUNCTION PROGRAM LIBRARY ENTRY EXIT INV CPU DB NON-DB WRITTEN WAITS
15 ð4 26 225
. ððððð1 RETURN QPTTRCJ1 QPFR ðð77 ððCF ð3 .ð12 1
15 ð4 26 262
. ððððð2 CALL QCLRTNE QSYS ððð1 ðð2D ð4 1
15 ð4 26 296
. ððððð3 XCTL QCLCLNUP QSYS ððð1 ðð48 ð4 .ð12
15 ð4 26 3ð7
. ððððð4 RETURN QPTTRCJ1 QPFR ððDð ððDð ð3 .ðð8
15 ð4 26 316
. ððððð5 RETURN QCMD QSYS ð16C ð153 ð2 .ð12
15 ð4 26 33ð
. ððððð6 CALL QMHRCMSS QSYS ððð1 ð37E ð3 .ð12 1
15 ð4 26 363
. ððððð7 CALL QMHGSD QSYS ððð1 ððF5 ð4 .ð12
15 ð4 26 372
. ððððð8 CALL QMHRTMSS QSYS ððð1 ð136 ð5 .ðð8
15 ð4 26 383
. ððððð9 RETURN QMHGSD QSYS ððF6 ð397 ð4 .ð16
15 ð4 26 397
. ðððð1ð CALL QWSPUT QSYS ððð1 ð8A6 ð5 .ð28
15 ð4 26 429
. ðððð11 XCTL QWSGET QSYS ððð1 ð27E ð5 .ð12
15 ð4 26 44ð
. ðððð12 CALL QT3REQIO QSYS ððð1 ðð55 ð6 .ð61 5 3 1
15 ð4 26.445 ðððð13 T3-ENTRY
15 ð4 26.447 ðððð14 T3REQIO-REQIO
15 ð4 31.285 ðððð15 T3DEQ-DEQ

Figure 10-3. Trace Job Information Report

FUNCTION This causes the trace entry to be recorded. The possible trace entries
are as follows:
Trace Entry Description
Call Call external.
Data A data trace.
Event Event handler.
EXTXHINV External exception handler.
EXTXHRET Call termination because of a return from an exception.
INTXHINV Internal exception handler.
INTXHRET Return from an exception.
INVEXIT Call because of a call exit routine.
ITERM Intervening call termination.
ITRMXRSG Call termination because of a resignaling exception.
PTRMTPP Process termination.
PTRMUNHX Termination because of an unhandled exception.
Return Return external.
RSMTRC Trace resumed.
SSPTRC Trace suspended.
XCTL Transfer control.
PROGRAM The name of the program for the entry.
LIBRARY The library name that contains the program associated with the trace
entry.

10-8 Performance Tools V4R2

 ENTRY The instruction in the program where the program was given control.
This is true when a program is nonobservant and observant.
EXIT The instruction number in the program where the program gave up
control.
INV The call level of the program.
CPU The approximation of the CPU used on this trace entry. This is a calcu-
lated value based on the time used and the CPU model being run.
DB The number of physical database reads that occurred for the entry.
NON-DB The number of physical nondatabase reads that occurred for the entry.
WRITTEN The number of physical writes that occurred for the entry.
WAITS The number of waits that occurred for the entry.

The read and write counts do not include any asynchronous I/O operations. The
counts indicate the number of I/O requests (either single or multiple page) sent to
the device, and describe the request queuing at the device.

Analyzing the Relationship of Programs and Database Files

Use the Analyze Program (ANZPGM) command and the Analyze Database File
(ANZDBF) command to print an overview of the programs and files used in an
application. The commands provide reports showing program-to-file use and phys-
ical and logical file relationships in the libraries.

Use the Analyze Database File Keys (ANZDBFKEY) command to print an overview
of the key structure of logical files in an application.

These commands provide you with a file and program use overview and key defi-
nition detail. It may be that your files or programs have changed since they were
first written and the file use has changed. For example, there may now be more
logical files over your physical files than the application currently needs. This situ-
ation can cause performance degradation, especially if many key field changes or
record adds occur. Remove any unneeded logical views.

Although you may use these commands infrequently, it is recommended that you
use them periodically to get a good understanding of the program-to-file relation-
ships and of the logical file structure used in the applications.

Analyze Program (ANZPGM) Command

Use the ANZPGM command to produce reports showing program-to-file and file-to-
program relationships.

When you use the ANZPGM command, the following printer files are created as
output:
File Description
QPPTANZP The program-to-file relationship report (Program-to-File Cross-
Reference)
QPPTANZP The file-to-program relationship report (File-to-Program Cross-
Reference)

Chapter 10. Performance Utilities—Manager Feature 10-9

 Figure 10-4 on page 10-10 shows an example of the Program-to-File Cross-
Reference Report.
12/ð1/95 13:37:ð9 Program to File Cross-Reference Page 1
File Usage
Record 1=In 2=Out
Library Program Program Text Description Object Library Format 4=Upd 8=?
---------- ---------- -------------------------------------------------- ----------- ----------- ---------- ----------
QPFR OLDPTCHGJR QAPMDMPT \LIBL 1
QDPTJTYP QPFR SFL 3
QDPTJTYP QPFR SFLCTL 3
QDPTJTYP QPFR QDPTF1 3
QDPTJTYP QPFR QDPTF2 3
QDPTJTYP QPFR QDPTF3 3
QDPTJTYP QPFR HELP1 3
QTRIDX QPFRDATA IDXREC 6
QJTYP1 QPFRTEMP IDXREC 1
QJTYP2 QPFRTEMP IDXREC 6
OLDPTTNSRP \FILE 8
QAPMJOBS 8
QSYSPRT 8
QTRIDX 8
QTRINTD 8
QTRINTU 8
QTRJOBI 8
QTRJOBO 8
QAPMDMPT &LIB 8
QAPMJOBS &LIB 8
QTRIDX &LIB 8
QTRJSUM &LIB 8
QTRTSUM &LIB 8
QDDSSRC \LIBL 1
QPTMPLST \LIBL
QPTTRIDX \LIBL
QDDSSRC QPFR 1
&TRCJOBS QTEMP 8

Figure 10-4. ANZPGM Program-to-File Cross-Reference Report

The ANZPGM Program-to-File Cross-Reference Report shows the following

columns:
Library and Program The name of the program that uses the file shown.
Program Text Description The program’s text description, if it was provided at
program creation.
Object and Library The name of the object that the program refers to, and the
name of the library the object is in.
Record Format The name of the formats in the file used by the program in the file
being referred to.
File Usage The manner in which the file is used by the program (1—input,
2—output, 4—update, 8—unknown, or any of the OR’d combinations of
these, such as 3—input-output, 6—output-update). For more informa-
tion and other values for the Display Database Relations (DSPDBR)
command and the Display Program References (DSPPGMREF)
command, see the CL Reference .

Figure 10-5 on page 10-11 shows an example of the ANZPGM File-to-Program

Cross-Reference Report.

10-10 Performance Tools V4R2

12/ð1/95 13:37:15 File to Program Cross Reference Page 1
File Usage
Record 1=In 2=Out
Library Object Format Library Program Program Text Description 4=Upd 8=?
----------- ----------- ---------- ---------- ---------- -------------------------------------------------- ----------
QMNADDTO
QMNGOMNU
QMNMAIN
QPTBATCH
QPTCPTRP
QPTCPTSL
QPTCPTWK
QPTLCKQ
QPTPGMX2
QPTSLECT
QPTSYSRP
QPTSYSSL
QPTSYSWK
QPTTRCJ1 STRJOBTRC CPP 2
QPTTRIDX
QPTTST1
QPTSYSRP
\FILE QPFR OLDPTTNSRP 8
\FILE QPTTNSRB 8
\FILE QPTTNSRP 8
\FILE QPFRTEMP 8
\NONE QPFR QPTTRCJð ENDJOBTRC CPP
\NONE QPTTRCJ1 STRJOBTRC CPP
QAJOBTRC QPTTRCRP ENDJOBTRC CPP 8
QAPMDMPT QPTCHGJT 8

Figure 10-5. ANZPGM File-to-Program Cross-Reference Report

The ANZPGM File-to-Program Cross-Reference Report shows the following

columns:
Library and Object The name and library the file is in.
Record Format The names of the record formats in the file.
Library and Program The names and library of the programs that use the file.
Program Text Description The program text description.
File Usage The manner in which the file is used (1—input, 2—output, 4—update,
8—unknown, and OR’d combinations of these values).

Analyze Database File (ANZDBF) Command

Use the ANZDBF command to print reports detailing physical and logical file
relationships.

When you use the ANZDBF command, the following files are created as output:
File Description
QPPTANZD The printer file that has the physical-to-logical database file relation-
ships report (Database Relation Cross-Reference).
QPPTANZD The printer file that has the logical-to-physical database file relation-
ships report (Logical File Listing).
QAPTAZDR The database file that serves as input to the ANZDBFKEY command.

Figure 10-6 on page 10-12 shows an example of the ANZDBF Database Relation
Cross-Reference Report.

Chapter 10. Performance Utilities—Manager Feature 10-11

| 12/ð1/95 14:29:31 Database Relation Cross Reference Page 1
| Type Depncy
| P=Phy Depnd Dependent Dependent Type
| L=Lgl File Library Count File Library D/A
| ----- ---------- ---------- ----- ---------- ---------- ------
| L QANSCRAL QSMU
| QANSCRA1
| QANSCRA2
| QANSCRA3
| QANSCRCL
| QANSCRL
| QANSCRML
| QANSCRM1
| QANSCRM2
| QANSCRM3
| QANSCRNL
| QANSCRN1
| QANSCR1
| QASVNUP
| P QANSCRAC 4 QANSCRAL QSMU D
| 4 QANSCRA1 QSMU D
| 4 QANSCRA2 QSMU D
| 4 QANSCRA3 QSMU D
| QANSCRAN 2 QANSCRNL QSMU D
| 2 QANSCRN1 QSMU D
| QANSCRCN 2 QANSCRCL QSMU D
| 2 QANSCRC1 QSMU D
| QANSCRCR 2 QANSCRL QSMU D
| 2 QANSCR1 QSMU D
| QANSCRMS 4 QANSCRML QSMU D
| 4 QANSCRM1 QSMU D
| 4 QANSCRM2 QSMU D
| 4 QANSCRM3 QSMU D
| QANSSRC
| QANSSRI
| QASVNUPP 1 QASVNUP QSMU D
| 32 records processed

Figure 10-6. ANZDBF Database Relation Cross-Reference Report

The ANZDBF Database Relation Cross-Reference Report has the following

columns:
Type The file type (P-Physical, L-Logical).
File The name of the file.
Library The library containing the file.
Depnd Count The number of logical files dependent on this file.
Dependent File The names of each dependent logical file.
Dependent Library The library the dependent logical files are in.
Depncy Type D/A D—Data share dependency. A—Access share dependency.

The entries in the Type, File, and Library columns are left blank if they are the
same as the previous line.

Figure 10-7 on page 10-13 shows an example of the ANZDBF Logical File Report.

10-12 Performance Tools V4R2

| 12/ð1/95 14:29:34 Logical File Listing Page 1
| Depncy Type
| Dependent Dependent Type P=Phy
| File Library D/A File Library L=Lgl
| ---------- ---------- ------ ---------- ---------- -----
| QANSCRAL QSMU D QANSCRAC QSMU P
| QANSCRA1 QSMU D
| QANSCRA2 QSMU D
| QANSCRA3 QSMU D
| QANSCRNL QSMU D QANSCRAN QSMU P
| QANSCRN1 QSMU D
| QANSCRCL QSMU D QANSCRCN QSMU P
| QANSCRC1 QSMU D
| QANSCRL QSMU D QANSCRCR QSMU P
| QANSCR1 QSMU D
| QANSCRML QSMU D QANSCRMS QSMU P
| QANSCRM1 QSMU D
| QANSCRM2 QSMU D
| QANSCRM3 QSMU D
| QASVNUP QSMU D QASVNUPP QSMU P
| 15 records processed
Figure 10-7. ANZDBF Logical File Report

The ANZDBF Logical File Report shows the following:

Dependent File The names of each dependent logical file.
Dependent Library The library the dependent logical files are in.
Depncy Type D/A D—Data share dependency. A—Access share dependency.
File The name of the physical file.
Library The library containing the physical file.
Type The physical file type.

Analyze Database File Keys (ANZDBFKEY) Command

Use the ANZDBFKEY command to print a report showing the key structure of
logical files.

When you use the ANZDBFKEY command, the following input file is used:

File Description
QAPTAZDR Database file that is the output from the ANZDBF command.
Note: Because the ANZDBFKEY command uses the output from the ANZDBF
command as its input, be sure the ANZDBF command is finished before
you use the ANZDBFKEY command. The ANZDBFKEY command tests the
existence of the ANZDBF output file and, if the file does not exist, the
program ends.

When you use the ANZDBFKEY command, the following files are created as
output:
File Description
QPPTANZK Printer file for the access path and record selection report (Key Fields
and Select/Omit Listing).

Chapter 10. Performance Utilities—Manager Feature 10-13

 QPPTANKM Printer file for the logical file key report (Analysis of Keys for Database
Files).

The information provided in these reports may suggest ways of combining logical
files, for physical files with a number of logical files over them. This process of
combining reduces the total number of logical files the system must maintain.

For example, consider an application that uses these two logical views of the same
physical file:
Ÿ Logical file FILEA with key FIELD1
Ÿ Logical file FILEB with keys FIELD1 and FIELD2

In this case, it is likely that you could delete FILEA and use FILEB instead.

Reducing the number of logical views an application uses can help the performance
of the application and of the system.

Figure 10-8 gives an example of the ANZDBFKEY Key Fields and Select/Omit
Listing.

This report lists the access path and selection (logical files only) values based on
the output produced by the Display File Description (DSPFD) command with a
single line for each key field or selection rule.

12/ð1/95 14:35:ð2 Key Fields and Select/Omit Listing Page 1

File Library Order Path Type Unique Maintenance
PHY QAOFCP QOFCFLS FIFO KEYED N \IMMED
Based on Format Key Field Seq Sign Zone Alt
NAME
JDATE SIGN
STIME SIGN
SEQ SIGN
EXT SIGN
GMTGNO SIGN
File Library Order Path Type Unique Maintenance
LGL QAOFCALL QOFCFLS FIFO KEYED N \IMMED
Based on Format Key Field Seq Sign Zone Alt
QAOFCP QOFCFLS CALRC1 MJDATE SIGN
MTIME SIGN
NAME
\\Record Selection\\ Format Field S/O Comp Values
CALRC1 MJDATE S GT +ð
MTIME A GT +ð
EXT A LE +2
O AL
Figure 10-8. ANZDBFKEY Key Fields and Select/Omit Listing

In the ANZDBFKEY Key Fields and Select/Omit Listing Report, the first output line
shows the following:
File The file name and, to the left of the name, the file type—physical (PHY)
or logical (LGL).
Library The name of the library in which the file is contained.
Order Ascending or descending sequence for the keys (LIFO, FIFO).
Path Type The type of access path (ARRIVAL, KEYED, or SHARED).

10-14 Performance Tools V4R2

Unique Whether unique keys are used (Y or N).
Maintenance *IMMED, *RBLD, or *DLY.

The second output line shows the following:

Based On The physical file name.
Format The format name in the logical file.
Key Field The name of the key field (can be one or more lines).
Seq The key sequence (blank is ascending, DES is descending).
Sign The key sign (blank, SIGN, or ABSV).
Zone The zone/digit specified (blank, ZONE, or DIGIT).
Alt The alternative collating sequence (YES or blank).

If record selection is used, the third output line shows the following:
Format The logical file format name.
Field The select/omit field name.
S/O Whether to select (S) or omit (O).
Comp The compare relation such as EQ, GT, LT, and AL (all).
Values The values to compare against.

Printer File QPPTANKM lists the file names, and for logical files, the key fields for
each format in descending order from major key to minor key.

You can use this list to find ways to combine logical files, when physical files have
many logical files over them. By combining files, you can reduce the number of
logical views an application requires and the total number of logical files the system
must maintain. Having fewer files to maintain can improve the performance of the
application and of the system.

Figure 10-9 on page 10-16 shows an example of the ANZDBFKEY Analysis of

Keys for Database Files Report.

The columns in the ANZDBFKEY Analysis of Keys for Database Files Report are
as follows:
Physical File The name of the physical file.
Library The physical file library.
File The logical files over the physical file.
Library The library the file is in.
Logical Format The logical file format name.
Maint Maintenance. Specify I (immediate), R (rebuild), or D (delay).
Key Fields Major to Minor Up to seven key fields.
No. Keys The number of key fields in the file.
S/O Whether select/omit is specified for key. YES indicates it is specified.

Chapter 10. Performance Utilities—Manager Feature 10-15

12/ð1/95 14:35:ð2 Analysis of Keys for Database Files Page 1
Physical File QAOFCP Library QOFCFLS
Logical Maint No.
File Library Format \\\\\\\ Key Fields Major to Minor \\\\\\\ Keys S/O
QAOFCP QOFCFLS I NAME JDATE STIME SEQ EXT GMTGNO 6
QAOFCALL QOFCFLS CALRC1 I MJDATE MTIME NAME 3 YES
QAOFCL QOFCFLS MTGREC I GMTGNO NAME JDATE STIME SEQ EXT 6

Figure 10-9. ANZDBFKEY Analysis of Keys for Database Files Report

Analyzing Process Information

Use the Display and Analyze Access Group (DSPACCGRP and ANZACCGRP)
commands to examine and analyze the process access group (PAG).

Use DSPACCGRP to report on PAG data for selected jobs. Use ANZACCGRP to
further analyze the DSPACCGRP output.

Process access group analysis provides you with a view of the operational environ-
ment for all jobs, or a group of jobs, in the system at a given time. Use the infor-
mation from process analysis to tune your system. When you tune your system,
you improve the program environment, causing a reduction in the number of the
following:
Ÿ Open files
Ÿ File buffer and work space sizes
Ÿ File open placement in a program

Display Access Group (DSPACCGRP) Command

Use the DSPACCGRP command to see the following:
Ÿ Size of a job’s PAG
Ÿ Open files
Ÿ I/O count for all files

When you use the DSPACCGRP command, the following output files are created:
File Description
QPPTPAGD Printer file
QAPTPAGD Database file (input to the ANZACCGRP command)

The following DSPACCGRP display appears only when you specify OUTPUT(*).

10-16 Performance Tools V4R2

à ð
6/16/96 Process Access Group Display DSPACCGRP
1ð:36:45 Sample Report
Job: QFILESYS1 User: QSYS Number: 144221 Type: S
PAG Size: 7ð4,512 Files: ð I/O count: ð

FILE LIBRARY MEMBER SIZE I/O COUNT

(No files open for this PAG)

F3=Exit F5=Refresh PRINT=QPPTPAGD

á ñ
When the PAG information is being displayed, the function key options to
DSPACCGRP are as follows:
1. Press F3 (Exit) or the Enter key to end the program.
2. Press F5 (Refresh) to display the PAG again to show any changes that
occurred since the current display was shown. If the job cannot be displayed
because its structure is changing, or the job ended, a message appears and
the command ends.
3. Press the Print key to print the displayed job’s PAG data. The format is the
same as the format used when you select OUTPUT(*PRINT).

Figure 10-10 shows an example of the DSPACCGR Process Access Group Infor-
mation Report. The fields on this report are the same as those on the display.

12/ð1/95 Process Access Group Information Page 1

15:ð1:59
Job: BYSINN User: VLLXR239 Number: ðð3368 Type: I
PAG Size: 164,352
FILE LIBRARY MEMBER SIZE I/O COUNT
---------- ---------- ---------- -------- ---------
QDGENDSP QSYS 9216 68
QRZLHEL3 QSYS QRZLHEL3 3ð72 1
QDPTPAGD QPFR 4ð96 3
Number of files: 3 Total I/O count: 72
Figure 10-10. Process Access Group Information

When you use the DSPACCGRP command, be aware of the following:

1. If the job you display is changing (such as calling and returning from programs
or opening and closing files), DSPACCGRP may try again several times to get
an accurate view of the PAG. DSPACCGRP tries five times before it stops and
displays a message telling you to try the command again.

Chapter 10. Performance Utilities—Manager Feature 10-17

 2. The command shows up to 999 file names and then stops. It tells you that
more files exist, but could not be displayed.
3. The I/O count (shown for each file) might give you some unexpected results.
For example, if you use subfiles on a display, several PUTs of records to the
subfile are included in the I/O count. This is in addition to the I/O operation
against the subfile control record that causes the subfile to display.
4. PAG SIZE is the total in-use size of the PAG. Because there can be a number
of inactive programs (programs that have returned), the total allocated size of
the PAG can be larger than the size currently in use. For example, a job that
was previously using interactive debug, but is now doing some other function,
could have an allocated size much larger than the in-use size.
5. To reduce the amount of main storage needed for the job PAG, group files that
are most frequently used. To do this, order the specification (or OPENs, if
explicit open is used) of files based on their I/O count.
DSPACCGRP shows the files in the order the system opens them (the RPG
implicit OPEN opens the files in reverse order from their order in the program).
It does not matter whether the most frequently used files are specified or
explicitly opened first or last, as long as they are grouped together by fre-
quency of use.

Analyze Process Access Group (ANZACCGRP) Command

Use the ANZACCGRP command to determine whether:
Ÿ Application programs use shared display and database files
Ÿ Files are ordered by their frequency of use
Ÿ Files are remaining open but have no activity

All of these conditions can affect system performance, especially if the jobs have
very large PAGs.

Consider the size of the PAG when you decide whether to use the PURGE(*NO)
job attribute when running an application. An application that requires fast response
time and has frequent and steady operator interaction would be a candidate for
running in a PURGE(*NO) environment.

The availability of main storage to dedicate to this application is the critical factor in
the decision to use PURGE(*NO). In general, the larger the PAG, the larger the
amount of main storage that would need to be dedicated.

By using the DSPACCGRP command job selection, the ANZACCGRP command

allows you to analyze the PAG of a single job, a set of jobs, or all jobs in the
system. It is most useful in examining a large number of PAGs collected using the
Job(*All), Job(*Int), or Job(generic name) options of the DSPACCGRP command.

The ANZACCGRP command uses the QAPTPAGD database file (the output from
the DSPACCGRP command) as input.
Note: Because the ANZACCGRP command uses the output from the
DSPACCGRP command, be sure to run the DSPACCGRP command first
with parameter OUTPUT(*FILE) or OUTPUT(*BOTH).

10-18 Performance Tools V4R2

 File Description
QSYSPRT Printer file

The ANZACCGRP command produces a three-part report:

Ÿ Environment Summary
Ÿ Job Summary
Ÿ File Summary

Figure 10-11 shows an example of the ANZACCGRP Environment

Summary. This section shows information for four categories of job types.

ANZACCGRP SUMMARY REPORT

Sample Report
DATA FROM: QPFRDATA /QAPTPAGD MBR: QAPAGDTA
DATE : 12/14/95 TIME : ð9:58:57
E N V I R O N M E N T S U M M A R Y
----------------------------------------
NUMBER AVERAGE # AVG # DUP AVG # DSP AVERAGE IO AVERAGE
JOB TYPE OF JOBS FILES/JOB FILES/JOB FILES/JOB COUNT/JOB PAG SIZE
------------------ ------- --------- --------- --------- ---------- --------
INTERACTIVE 4 2 2 44 2ð7872
BATCH/AUTOSTART 4 188 878ð8
READER/WRITER 1 3 35 126464
OTHER 6 5 75 135253
----- ----- ----- ---- ----------- --------
TOTAL/AVERAGE 15 3 2 94 14138ð
Figure 10-11. Environment Summary

The following columns appear on the Environment Summary:

JOB TYPE The type of jobs the summary line is for (Interactive,
Batch/Autostart, Reader/Writer, or Other).
NUMBER OF JOBS The number of jobs of the corresponding type.
AVERAGE # FILES/JOB The average number of open files.
AVG # DUP FILES/JOB The average number of files open more than once
for each job. If this number is high, it indicates that files are
being opened with SHARE(*NO).
AVG # DSP FILES/JOB The average number of open display files for each
interactive job. If this number is high, it indicates that display files
are defined SHARE(*NO) or that several display files are used
by an application. If there are not a large number of display file
formats that take a lot of PAG space, consider using
SHARE(*YES) and combining display files.
AVERAGE I/O COUNT/JOB The average number of file operations for all
open files, not opened since the job started.
AVERAGE PAG SIZE The average PAG size for each job type.

Figure 10-12 shows an example of the ANZACCGRP Job Summary. This

section shows information for all selected jobs.

Chapter 10. Performance Utilities—Manager Feature 10-19

 ANZACCGRP SUMMARY REPORT

DATA FROM: QPFRDATA /QAPTPAGD MBR: QAPAGDTA

DATE : ð6/14/96 TIME : ð9:58:2ð
J O B S U M M A R Y
-----------------------
TOTAL DUP. DSP. SIZE OF SIZE OF MULT
JOB FILES FILE FILE TOTAL I/O SIZE OF DUP DSP ACT PGM
JOB NAME USER NAME NUMBER TP PAG SIZE USED USED USED COUNT FILES FILES FILES PGM CALL
---------- ---------- ------ -- ----------- ----- ---- ---- ------------- ----------- ----------- ----------- --- ----
QFILESYS1 QSYS 144221 S 7ð4512
QDBSRVXR QSYS 144222 S 421888 19 159 155648
Q4ððFILSVR QSYS 144223 S 516ð96
QQQTEMP1 QSYS 144224 S 258ð48
QQQTEMP2 QSYS 144225 S 253952
QDBSRVXR2 QSYS 144226 S 36ð448 7 2 4ð9 942ð8 53248
QSYSCOMM1 QSYS 144227 S 26624ð
QPADEVððð9 SOFIACN 144592 I 438272 2 2 29 28672 28672
QPADEVðð16 SOFIACN 144593 I 692224 7 2 35 65536 24576
QPADEVðð17 SOFIACN 144594 I 4ð14ð8 3 2 564 32768 28672
QPADEVðð15 LUPITA 144596 I 36864ð 1 1 3 2ð48ð 2ð48ð
QCTL QSYS 144229 M 9ð112ð
QSYSWRK QSYS 14423ð M 1ðð7616
QSTRUPJD QPGMR 144231 A 4ð96ðð
QSPL QSYS 144239 M 831488
QSERVER QSYS 14424ð M 999424
QPADEVððð3 REYNAGAG 1443ð3 I 413696 1 1 8 2ð48ð 2ð48ð
QZDAINIT QUSER 144242 B 1466368 4 32768
QSYSSCD QPGMR 144243 B 282624
QPWFSERVSO QUSER 144244 B 831488
QINTER QSYS 144245 M 888832
QSERVER QPGMR 144246 A 1134592

Figure 10-12. ANZACCGRP Job Summary

The Job Summary contains the following columns:

JOB NAME/USER NAME/JOB NUMBER The name of the job, name of the
user, and job numbers.
TP The type of job. The various types of codes and descriptions
follow:

Job Type Code Job Type Description

A Autostart
B Batch
I Interactive
M Other
R Reader
S Other
W Writer
X Other
PAG SIZE The gross size of the PAG for the job.
TOTAL FILES USED The number of all open files for the job (including
display and spooled files).

10-20 Performance Tools V4R2

DUP. FILE USED The number of files with more than one ODP in the PAG.
If this number is not zero, it indicates that files are being opened
SHARE(*NO). If possible, the files should be combined and
created with SHARE(*YES).
DSP. FILE USED The number of display files open for the job. If this
number is high, it indicates that the display files are being
opened SHARE(*NO) or that the application has multiple display
files. If possible, display files should be combined and created
with SHARE(*YES).
TOTAL I/O COUNT The count of file operations (including OPEN and
CLOSE) for all open files in the job. This, however, is not the
total I/O count for the job.
SIZE OF FILES The amount of space in the PAG that is occupied by ODPs
for all open files.
SIZE OF DUP FILES The amount of space in the PAG that is occupied by
ODPs for files opened with SHARE(*NO). The first ODP for each
duplicate file is not included in this total. This value represents
the amount of space you will save if you use SHARE(*YES).
SIZE OF DSP FILES The amount of space in the PAG that is occupied by
ODPs for display files.
ACT PGM The number of programs in the call stack that are active.
MULT PGM CALL The number of active programs that appear more than
once in the call stack.

Figure 10-13 shows an example of the ANZACCGRP File Summary. This

section shows information for all open files.

Chapter 10. Performance Utilities—Manager Feature 10-21

 ANZACCGRP SUMMARY REPORT
Sample Report
DATA FROM: QPFRDATA /QAPTPAGD MBR: QAPAGDTA
DATE : 12/14/95 TIME : ð9:58:57
F I L E S U M M A R Y
-------------------------
FILE NUMBER NUMBER OF TOTAL AVERAGE
FILE NAME LIBRARY MEMBER TYPE OF JOBS TIMES OPEN I/O COUNT ODP SIZE
----------- ---------- ---------- ---- ------- ---------- ----------- --------
QRZPHELM QSYS QRZPHELM DB 1 1 256ð
QRZPHVPD QSYS QRZPHVPD DB 1 1 8 256ð
QRZPHDEL QSYS QRZPHDEL DB 1 1 256ð
QRZPHLOC QSYS QRZPHLOC DB 1 1 256ð
QRZPPTF QSYS QRZPPTF DB 1 1 256ð
QRZPLCG QSYS QRZPLCG DB 1 1 256ð
QRZPTCBL QSYS QRZPTCBL DB 1 1 256ð
QRZPTOEM QSYS QRZPTOEM DB 1 1 256ð
QRZPTSPC QSYS QRZPTSPC DB 1 1 256ð
QRZLHEL1 QSYS QRZLHEL1 DB 1 1 135 3ð72
QRZLHEL2 QSYS QRZLHEL2 DB 1 1 3ð72
QRZLHEL3 QSYS QRZLHEL3 DB 2 2 246 3ð72
QRZLHEL4 QSYS QRZLHEL4 DB 1 1 8 3ð72
QRZLHEL5 QSYS QRZLHEL5 DB 1 1 3ð72
QRZLHEL7 QSYS QRZLHEL7 DB 1 1 3ð72
QRZLHVP1 QSYS QRZLHVP1 DB 1 1 59 256ð
QRZLHVP2 QSYS QRZLHVP2 DB 1 1 3ð72
QRZLLOC1 QSYS QRZLLOC1 DB 1 1 256ð
QRZLTEL1 QSYS QRZLTEL1 DB 1 1 3ð72
QRZLTEL2 QSYS QRZLTEL2 DB 1 1 3ð72
QRZLHEDR QSYS QRZLHEDR DB 1 1 3ð72
QRZLDSUA QSYS QRZLDSUA DB 1 1 3ð72
QRZLDSU1 QSYS QRZLDSU1 DB 1 1 3ð72
QRZLPATH QSYS QRZLPATH DB 1 1 256ð
QRZLPTH1 QSYS QRZLPTH1 DB 1 1 3ð72
QRZLLCG1 QSYS QRZLLCG1 DB 1 1 3ð72
QRZLLCG2 QSYS QRZLLCG2 DB 1 1 256ð
QADBXREF QSYS QADBXREF DB 1 1 256ð
QADBFDEP QSYS QADBFDEP DB 1 1 256ð
QADBLDEP QSYS QADBLDEP DB 1 1 256ð
QADBLDNC QSYS QADBLDNC DB 1 1 3ð72
QADBXDIC QSYS QADBXDIC DB 1 1 256ð
QADBXFIL QSYS QADBXFIL DB 1 1 3ð72
QADBXOWN QSYS QADBXOWN DB 1 1 3ð72
QPSPLPRT QSYS DS 1 1 12 46ð8
Qð4ð79Nðð1 QSPL Q979578132 DB 1 1 2ð 256ð
QDSPUIM QSYS DS 2 4\ 67 19968
QDGENDSP QSYS DS 2 2 46 16384
QDDSPEXT QSYS DS 1 1 8 46ð8
QAPTPAGD QPFRDATA QAPAGDTA DB 1 1 57 256ð
NETFILE QNET DS 1 1 745 3584
NET38SPOOL QNETSPOOL VN6A633564 DB 1 1 7 13824
Figure 10-13. ANZACCGRP File Summary

The ANZACCGRP File Summary contains the following columns:

FILE NAME/LIBRARY/MEMBER Name of the open file ODP.
FILE TYPE One of the following:
DB Database or spooled file
DS Display or ICF file
SP Spooled file being created

10-22 Performance Tools V4R2

NUMBER OF JOBS The total number of jobs using the file.
NUMBER OF TIMES OPEN The total number of times the file has been
opened. If this number is larger than the number of jobs (indi-
cated by an *), the file has been created with SHARE(*NO).
TOTAL I/O COUNT The total number of file operations (including OPEN
and CLOSE) that have been performed by all jobs for currently
open files. If this number is very low, the applications should be
changed to open the file only when needed.
AVERAGE ODP SIZE The average amount of space in the PAG occupied
by each open occurrence of the file. In general, this number
represents the ODP size for the file.

Chapter 10. Performance Utilities—Manager Feature 10-23

10-24 Performance Tools V4R2
 Chapter 11. Performance Explorer
Performance explorer is a data collection tool that helps the user identify the
causes of performance problems that cannot be identified by collecting data using
the performance monitor or by doing general trend analysis. Two reasons to use
performance explorer include:
Ÿ Isolating performance problems
Ÿ Modeling the performance of applications

| The collection functions and related commands of performance explorer are part of
| the OS/400 operating system. The reporting function and its associated commands
| are part of the Performance Tools for AS/400 licensed product, the Manager
| feature. The AS/400 Performance Explorer Tips and Techniques, SG24-4781, pro-
| vides additional examples of the performance explorer functions and examples of
| the enhanced performance explorer trace support.

Do I Need Performance Explorer?

Performance explorer is a tool that helps find the causes of performance problems
that cannot be identified by using tools that do general performance monitoring. As
your computer environment grows both in size and in complexity, it is reasonable
for your performance analysis to gain in complexity as well. The performance
explorer addresses this growth in complexity by gathering data on complex per-
formance problems.

This tool is designed for application developers who are interested in understanding
or improving the performance of their programs. It is also useful for users know-
ledgeable in performance management to help identify and isolate complex per-
formance problems.
Note: If you are familiar with the Sampled Address Monitor (SAM) function or the
TPST PRPQ, your transition to the performance explorer should be smooth.

Who Needs Performance Explorer

The user that wants to use the performance explorer for application and program
analysis does not need to be an expert in the area of performance to make use of
the performance explorer. The same level of expertise that was required when
using SAM in V3R1 and earlier versions is what is needed with the performance
explorer.

To use the performance explorer to isolate a performance problem, you should

have a good understanding of the performance issue. Using the performance
explorer requires more performance specialized knowledge. You should also have
an idea of where the problem might be on the system. You need to set up perform-
ance explorer to collect data in specific areas of your system, and you will have to
interpret the data.

 Copyright IBM Corp. 1998 11-1

 When You Need Performance Explorer
When you find that performance advisor is not telling you enough, you should con-
sider the performance explorer. In short, performance explorer is the tool you need
to use after you have tried the other tools. It gathers specific forms of data that can
more easily isolate the factors involved in a performance problem.

Comparison of Explorer to Other Performance Tools

A good way to understand performance explorer is to see it compared and con-
trasted to other tools in the Performance Tools licensed program or in the OS/400
operating system.

Performance Explorer and Advisor Functions

The performance advisor and the performance explorer are quite different functions.
The explorer's main purpose is collecting specific data. To do this, it has its own
collecting facility. The advisor's role is assessing data collected by the performance
monitor. It produces, after its analysis, a list of conclusions and recommendations
on ways you can improve your performance. The explorer does not do any analysis
for you.

If you are using the advisor, you are probably doing routine performance mainte-
nance. If you are using the explorer, you know that you have a performance
problem, and you are having a hard time identifying its cause.

Performance Explorer and Performance Monitor

In a sense, the performance explorer is much like the performance monitor
because they both collect data. The main difference is that performance explorer
provides a much greater level of detail. Also, unlike the performance monitor, the
performance explorer allows you to specify particular areas of interest, and it allows
you to focus the collection. The performance explorer collection can be tuned to
include very specific data. It is the ability to tune, or specify, the data to be col-
lected that makes the performance explorer effective in helping isolate performance
problems.
| Note: You can run both collections of data at the same time. However, you should
| keep this to a minimum because the system is significantly affected when
| both collections are active.

This detail is good because many times the factors that contribute to performance
problems are not isolated. Often, there are several contributing factors to the
problem. Since many factors are contributing to the problem, it is difficult to identify
individual causes using tools that assess the whole system. You need a tool that
provides the details.

Performance explorer can be used to isolate the factors behind a problem that you
have identified. This tool pulls the symptoms together. For example, a sluggish
system could be caused by DASD reads and writes, or CPU, or a combination of
two or more. With performance explorer, you can find the cause.

11-2 Performance Tools V4R2

Benefits of Performance Explorer
Performance explorer has advantages for people who need detailed performance
analysis on an AS/400 system. Using performance explorer you can:
Ÿ Do a detailed analysis on one job without affecting the performance of other
operations on the system.
Ÿ Analyze data on a system other than the one on which it was collected. For
example, if you collect data on a managed system in your network, you can
send it to the central site system for analysis.
Ÿ Map performance information to code
Using performance explorer, you can map performance information back to
source lines of code to correlate the performance data generated with the code
that caused the data to be generated.
Ÿ Collect performance information on user-developed software.
To collect performance information on user-developed software, performance
collection must be enabled when the program is created. In general, all user-
developed software is created with performance collection enabled.
For more information on how to enable or disable performance collection, refer
to the specific compiler documentation and refer to the Enable performance col-
lection (ENBPFRCOL) parameter on the Create Bound C Program (CRTBNDC)
command.
Programs can also be enabled or disabled using the ENBPFRCOL parameter
on the Change Program (CHGPGM) command.
Note: The default for all ILE languages is to have the pre-defined trace points
at the program-level enabled. However, some languages provide a com-
piler option (ENBPFRCOL parameter) that allows you to turn the ena-
bling off. Those languages that do not provide the option will have the
pre-defined collection points enabled.
The significance of the collection mechanism is that:
– It is controlled by pre-defined collection points that are compiler generated.
– The pre-defined collection points are scalable.
– The system and all IBM code are shipped with these pre-defined trace
points.
– The default for all compilers is to have these pre-defined collection points
enabled.

How Performance Explorer Works

1. You set up a performance explorer data collection using a definition.
2. You start performance explorer and it collects the data based on the definition.
3. You can create reports from the databases.
4. You can print those reports, if you want to.

You can access the commands associated with the performance explorer tool using
one of the following:

Chapter 11. Performance Explorer 11-3

 Ÿ The command interface. Type the commands from the command line. All the
commands are part of the OS/400 operating system, except the PRTPEXRPT
command.
Ÿ The Performance Tools menu options. Select option 5 (Performance utilities)
from the IBM Performance Tools menu, then option 2 (Work with Performance
Explorer).

à ð
Work with Performance Explorer

Select one of the following:

1. Add Performance Explorer Definition (ADDPEXDFN)

2. Change Performance Explorer Definition (CHGPEXDFN)
3. Remove Performance Explorer Definition (RMVPEXDFN)
4. Start the Performance Explorer (STRPEX)
5. End the Performance Explorer (ENDPEX)
6. Print Performance Explorer Reports (PRTPEXRPT)
7. Delete Performance Explorer Data (DLTPEXDTA)

Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ

Performance Explorer Definitions

The performance explorer definitions are how you describe to the performance
explorer the type of data that you want to collect. The definition is then used to
define the information that should be collected when the performance explorer is
started. When you create a performance explorer definition, you are really just
describing what you want collected. To help you understand how performance
explorer works, scenarios are provided to help you understand the concepts of per-
formance explorer. See “Creating a Performance Explorer Definition” on page 11-6.

The performance explorer provides the following types of data collection:

Ÿ Statistical
Ÿ Profile
Ÿ Trace

Each type gathers data in a different way and organizes it in a unique fashion.

| Statistical type identifies applications and IBM programs or modules that consume
| excessive CPU use or that perform a high number of disk I/O operations. Typically,
| you use the statistical type to identify programs that should be investigated further
| as potential performance bottlenecks.

11-4 Performance Tools V4R2

| You use the statistical type to find the high resource consuming programs and
| modules that run during the performance collection. The objective is to determine if
| there are specific programs or modules consuming most of the resource. If you
| suspect that using a particular application is causing poor performance, use the
| *STATS collection. The *STATS collection can determine which programs should
| be examined based on the CPU statistics, disk statistics, and the number of times
| the program was called.

| For example, you may find that the highest CPU utilization program or module is
| called only once during the collection period. You can have another program or
| module that uses a moderately high level of CPU utilization but is called hundreds
| of times. You need to decide which program or module should be analyzed: the
| one called only once or the one called hundreds of times.

The statistics can be structured in either a hierarchical or flattened manner.

Ÿ A hierarchical structure organizes the statistics into a call tree form in which
each node in the tree represents a program procedure run by the job or task.
Ÿ A flattened structure organizes the statistics into a simple list of programs or
procedures, each with its own set of statistics.

| Profile type identifies high-level language (HLL) programs that consume excessive
| CPU utilization based on source program statement numbers.

| You can also identify a program that is constantly branching between the start of
| the program and subroutines at the end of the program. If the program is large
| enough, this constant jumping back and forth can cause excessive page fault rates
| on a system with limited main storage.

When a trace, statistics, or profile mode is selected, a number of options are pro-
vided to limit the scope of the collected performance data. You can collect on spe-
cific jobs, tasks, OS/400 programs, modules, and procedures.

Trace type gathers a historical trace of performance activity generated by one or

more jobs on the system. The trace type gathers very specific information about
when and in what order events occurred. The trace type collects detailed program,
Licensed Internal Code (LIC) task, OS/400 job, and object reference information.

General Flow of the Performance Explorer

The following sections should help you become familiar with the natural path
through the performance explorer. Figure 11-1 on page 11-6 shows a basic work
cycle.

Chapter 11. Performance Explorer 11-5

 ADDPEXDFN QPEXDATA
Command Database

Database PRTPEXRPT
Command

Runs
STRPEX Collections ENDPEX
Command Command
RV3S161-0

Figure 11-1. Performance Explorer Basic Flow Pattern

The work cycle is made up of these activities:

| Ÿ The first task in this cycle is to create a session definition that informs the
| AS/400 system about what processes you want to collect performance data. On
| the Add Performance Explorer Definition (ADDPEXDFN) command, specify the
| collection type and a name for the definition. This definition is stored as a data-
| base member by that name in the QAPEXDFN file in library QUSRSYS. The
| name that you specify is used on the STRPEX command.
Ÿ The second task is to start collecting data (STRPEX command), which in turn
creates a data file containing the specified performance data.
Ÿ The third task is to stop collecting the data and save it to database files for
analysis. Use the End Performance Explorer (ENDPEX) command to stop the
collection.
Ÿ The fourth task is to analyze the performance data. The PRTPEXRPT
command provides unique reports for each type of data (statistical, profile, or
trace).
The other option for analysis is to write your own queries over the set of data-
base files.

Creating a Performance Explorer Definition

The first task is to define what data is to be collected using the Add Performance
Explorer Definition (ADDPEXDFN) command. After the definition is completed and
saved, you are ready to continue to the second task in the cycle of work.

Before creating a new definition, consider what kinds of information you want and
the amount of detail you need. In general, the three main types of collections have
the following characteristics:
Ÿ Statistics type definitions
– Using this definition results in collecting the same basic information as the
TPST tool.
– Good for first order analysis of OS/400 original program model (OPM) pro-
grams, procedures, and MI complex instructions.
- Gives number of invocations
- Gives both inline and cumulative CPU usage in microseconds

11-6 Performance Tools V4R2

 - Gives both inline and cumulative number of synchronous and asynchro-
nous I/O
- Gives number of calls made
– Works well for short or long runs
– Size of the collected data is fairly small and constant for all runs
– Run time collection overhead of ILE procedures may be a problem due to
the frequency of calls. Although run time is degraded, performance explorer
removes most of the collection overhead from the data.
– Uses combined or separated data areas. The MRGJOB parameter on the
Add Performance Explorer Definition (ADDPEXDFN) command specifies
whether all program statistics are accumulated in one data area, or kept
separate (for example, one data area for each job).
Ÿ Profile type definitions
– Gives detailed breakdown of where you are spending time within a program
or procedure
– Size of collection is fairly small and constant regardless of length of run
– Can narrow the scope of data collected to just a few programs of interest
– Limit of 16 MI programs means that you should use this as a second order
analysis tool.
– Can vary overhead by changing sample interval. An interval of 2 millisec-
onds seems a good first choice for benchmarks.
– No restrictions on pane size due to the number of programs specified or
the size of the programs specified.
Ÿ Trace type definitions
– Storage management and flow trace definitions
– Good for watching storage management activity on the system. Also shows
MI complex instructions.
– Longer runs collect more data

Starting the Performance Explorer

To start the performance explorer, use the Start Performance Explorer (STRPEX)
command. You can specify to start a new performance explorer session or resume
an already active session.
Note: You are allowed to have only one active session at a time. Multiple ses-
sions are not allowed.

Ending the Performance Explorer

To end the performance explorer session, use the End Performance Explorer
(ENDPEX) command. The ENDPEX command performs the following actions on
the collected data:
| Ÿ Places the collected data in files QAYPExxx in the library that you specify.

Chapter 11. Performance Explorer 11-7

| Use OPTION(*END) and DTAOPT(*LIB) to do this. The database member
| name for all the QAYPExxx files uses the session name as the default unless
| you specify a name for the DTAMBR parameter.
| You can specify RPLDTA(*NO) to erase data that was collected using this
| session name or RPLDTA(*YES) to add the collected data to the existing data.
| Unless you are a very sophisticated user, use RPLDTA(*NO).
| Ÿ Places the collected data into a single IBM-defined file.
| Use OPTION(*END) and DTAOPT(*FILE) to do this. Typically, you would use
| *FILE only under the direction of an IBM service representative. Specifying the
| *FILE value on the DTAOPT parameter saves the collection information into a
| binary file. The binary file option should be used only if the data is going to be
| shipped to IBM. The performance tools can analyze only the database files.
| Ÿ Discards the collected data.
| Use OPTION(*END) and DTAOPT(*DLT) to delete any collected data. You do
| this when you determine the collected data cannot be used. For example, one
| of the suspected jobs did not start as expected. If you choose the *DLT option,
| the collected performance data for the session is never saved.
| Ÿ Saves the collected data.
| Use OPTION(*END) and DTAOPT(*LIB) to save the collection into a database
| file. Use these values if you are sending data to a manager site.
| Ÿ Suspends the collection session but does not end it.
| Use OPTION(*SUSPEND) to do this. You can later start the data collection
| again by issuing the STRPEX command with OPTION(*RESUME) for the spe-
| cific session ID.
| Note: If you forget the active collection session name, you cannot end the
| collection without powering off the system. To find the active session
| name, use the ENDPEX SSNID(*SELECT) command.

Deleting Performance Explorer Data

To delete performance explorer data, use the Delete Performance Explorer Data
(DLTPEXDTA) command. The DLTPEXDTA command discards performance data
from a set of database files.

| Creating and Printing Performance Explorer Reports

You create and print performance explorer reports by using the Print Performance
Explorer Report (PRTPEXRPT) command.

| Use the OUTFILE parameter when you want to customize your Trace Report. The
| performance explorer stores its collected data in the QAVPETRCI file, which is
| located in the QPFR library. Type the following command to view the contents for a
| single record:
| DSPFFD FILE(QPFR/QAVPETRCI)

11-8 Performance Tools V4R2

 Finding Your Performance Explorer Definitions
The SELECT parameter on the ENDPEX command provides a list of all active per-
formance explorer sessions on the system. This parameter shows the Select Per-
formance Explorer Session display.

à ð
Select Performance Explorer Session

Type option, press Enter.

1=Select

Event
Option Session User Type State Count

(No objects to display.)

Types of Performance Explorer Reports

| Table 11-1 identifies the sections that are available for the individual performance
| explorer reports. Some sections are common to all reports; some reports have
| unique sections. The information that follows the table shows examples of each
| section.

| Table 11-1. Sections that are available for the performance explorer reports
| Section Statistics Report Profile Report Trace Report Base Report
| Definition X X X X
| Run X X X X
| Task X X X X
| CPU Summary X X
| Library X X
| Main X X X

Common Report Sections

The following series of report examples show the sections that are common to the
performance explorer reports. Each section also contains the field descriptions for
each report.

Definition Information
You define what kind of data to collect with the ADDPEXDFN command. The Defi-
nition Information report reflects the definition that was used in collecting the data.
This heading appears only once in any type of report.

| Figure 11-2 on page 11-10 shows an *STATS collection type as an example.

Chapter 11. Performance Explorer 11-9

| Performance Explorer Report
| Definition Information Page 1
| Library . . : QPEXDATA
| Member. . . : STATSF
| Description : BLANK
| Type . . . . . . . . . . . . : STATISTICS
| Definition Name. . . . . . . : SFSTATSF
| Defined By . . . . . . . . . : FOLEY
| Definition Description . . . : stats job(\) task(\none) dtaorg(\flat)
| Data Organization. . . . . . : FLAT
| Overhead Subtraction . . . . : YES
| Merge Jobs . . . . . . . . . : YES
| Include Dependent Jobs . . . : YES
| Selected Jobs:
| Name User Number
| \
| Selected Task Names:
| \NONE
| Selected MI Complex Instructions:
| \ALL

| Figure 11-2. Example for *STATS Definition Information

The Definition Information shows the following values:

| Library The library that contains the collected data.
| Member The name of the member that contains the collected data.
Description The description of the data that was saved.
Type The method for collecting performance data.
Definition Name The name of the performance explorer definition.
Defined By The user ID that created the definition
Definition Description The description given to the definition
Include Dependent Jobs The tasks or jobs that are created on behalf of a task or
job that is currently part of the collection are part of the collection.
Selected Jobs The jobs that are included in the performance explorer data col-
lection session. The “*” means the current job when the STRPEX
command is issued.
Name The name of the job to include in the performance explorer
data collection session. You can specify up to 10 qualified
jobs.
User The name of the job associated with a specific user. When
you specify a user name, you further qualify the job.
Number The number of the job. When you specify a job number, you
further qualify the job.
Selected Task Names The name of the Licensed Internal Code (LIC) low-level
task to be part of the collection. AS/400 Licensed Internal Code Diag-
nostic Aids - Volume 1 contains a list of LIC tasks and descriptions.
Note: By default all task names are included in each collection.
Selected MI Complex Instructions The machine interface (MI) complex
instructions that are part of the collection. The MI complex instructions
represent all the high-level machine interface instructions used by
OS/400 support. MI complex instructions include functions like finding

11-10 Performance Tools V4R2

 the pointer to an object, writing records sequentially to a file, or creating
a duplicate object.
Note: By default all MI instructions are included in each collection.
Sample Interval (ms) The rate, in milliseconds, that profile mode collections
sample the location for the currently running programs.
Selected Programs The programs listed as part of a profile type definition.
Pane Size The pane size is the number of consecutive program instruc-
tion addresses assigned to each counter. Pane size can
range from 4 bytes to 4096 bytes. The default pane size is 4
bytes. Valid values are 4, 8, 16, 32, 64, 128, 256, 512, 1024,
2048, and 4096. The smaller the pane size, the more
detailed the program profile information will be.
For example, a pane size of 4 means one instruction. A pane
size of 2048 means 512 instructions.
Library The library the program is in.
Program The program whose performance profile data will be col-
lected.
Type The type of program specified, either a program (*PGM)
object or a service program (*SRVPGM) object.
Module The specific module within the program or service program
that is to be profiled.
Procedure The specific procedure within the specified module that is to
be profiled.

Run Information
This report provides general information about when the data was collected, the
state of the machine from which the data was collected, details about the length of
collection, and who ran the collection. This heading appears only once in any type
of report.

Figure 11-3 on page 11-12 shows an example of the Run Information section. The
Run Information section provides the same information for each of the main reports,
which is general system and session information.

Chapter 11. Performance Explorer 11-11

| Performance Explorer Report
| Run Information Page 2
| Library . . : QPEXDATA
| Member. . . : STATSF
| Description : BLANK
| Sessions since IPL. . . . . . . . . . . . . : 1
| Session name. . . . . . . . . . . . . . . . : FOLEY
| Start time. . . . . . . . . . . . . . . . . : 1998-11-ð6-1ð.4ð.ð2.745ð8ð
| Stop time . . . . . . . . . . . . . . . . . : 1998-11-ð6-1ð.4ð.51.567576
| Total time DD-HH.MM.ss.uuuuuu. . . . . . . : ðð-ðð.ðð.48.822496
| Suspend time (us) . . . . . . . . . . . . . : 13,549,392
| Duration of trace . . . . . . . . . . . . . : 352731ð4
| Number of events. . . . . . . . . . . . . . : 2,332
| Trace wrap count. . . . . . . . . . . . . . : ð
| Missed events due to buffering. . . . . . . : ð
| Missed events while recording . . . . . . . : ð
| Job creating session. . . . . . . . . . . . : QPADEVðð29FOLEY 1ð1ð29
| Started by user . . . . . . . . . . . . . . : FOLEY
| Target system . . . . . . . . . . . . . . . : ABSYSTEM
| Serial number . . . . . . . . . . . . . . . : 1ð-18ð3D
| System type . . . . . . . . . . . . . . . . : 94ð6
| System model. . . . . . . . . . . . . . . . : 5ðð
| Total pages memory. . . . . . . . . . . . . : 32,768
| OS/4ðð level. . . . . . . . . . . . . . . . : 37ð
| Version . . . . . . . . . . . . . . . . . . : V4R2Mð
| Configured ASPs . . . . . . . . . . . . . . : 1
| Logical DASDs . . . . . . . . . . . . . . . : 4
| Jobs/tasks in session . . . . . . . . . . . : 1
| Jobs in session . . . . . . . . . . . . . . : 1

| Figure 11-3. *STATS Run Information Report

| The Run Information section shows the start, stop,and total run times of the col-
| lection. You also see that the job ran the collection on system ABSYSTEM.

The Run Information shows the following values:

Library The library the database file is in.
Member The database file member containing the data.
Description The description of the data that was saved.
Sessions since IPL The number of times that the performance explorer collected
data since the last IPL
Session name The name of the session.
Start time The time that the session was started
Stop time The time that the session was stopped.
Total time The total elapsed time that data was being collected, including sus-
pended time.
Suspend time (us) The amount of time, in microseconds, that the session was in a
suspended state. It's possible for this field to show a number other than
zero even if you have not suspended the session.
| Duration of trace (us) The total active runtime minus the total suspended time.
Number of events The total number of events encountered during a session.
Trace wrap count The trace wraps to the beginning when the trace file is full. The
oldest trace records are written over by new ones as they are collected.
This is the number of times the trace wrapped.

11-12 Performance Tools V4R2

 Notes:
1. Trace wrap count applies only to trace type session.
2. Specify the *STOPTRC value on the TRCFULL parameter of the
ADDPEXDFN command to avoid this wrapping.
| Missed events due to buffering The number of events that were not collected
| during the session.
| Missed events due to recording The number of events that were not collected
| during the session.
Job creating session The name of the job that issued the STRPEX command
Started by user The user ID that issued the STRPEX command
Target system The name of the system the data was collected on
Serial number The serial number of the system the data was collected on
System type The type of system the data was collected on
System model The model of system on which the data was collected
| Total pages memory The number of 4-K memory pages on the system on which
| the data was collected. 98,304 means 384MB of main storage.
| 1. 98,3ð4 / 1ð24(1K) = 96
| 2. 96 \ 4 = 384
OS/400 level The OS/400 driver level of the system on which data was collected
The OS/400 level relates to the latest level of cumulative package that is
installed on your system.
Version The OS/400 release level of the system on which data was collected
Configured ASPs The number of ASPs on the system on which data was col-
lected
Logical DASDs The physical number of disk drives attached to the system, if mir-
roring is not used. If your system has mirrored protection active, the mir-
rored disk pair count as one logical DASD.
Data areas The number of performance explorer internal data areas the collection
was partitioned into. If you specify MRGJOB(*YES), the performance
explorer stores the collected data for all jobs in a combined data area.
Jobs/tasks in session The total number of tasks and jobs that were included in
the session.
| Jobs in session The total number of jobs that were included in the session. In this
| example, this number is equal to Jobs/tasks in session minus Jobs in
| session. There is no correlation between jobs and LIC tasks. The col-
| lections always contain all the LIC tasks of the system.

Task Information
This report shows the jobs and task from which data was collected. AS/400
Licensed Internal Code Diagnostic Aids - Volume 1 contains a list of LIC tasks and
descriptions.

| Figure 11-4 on page 11-14 shows an example of the Task Information section. The
| Task Information section provides the same information for each of the main

Chapter 11. Performance Explorer 11-13

| reports. If the task or job did meaningful work during the collection time period, the
| values under the CPU (us) and CPU Percent are a number other than 0.

| Performance Explorer Report

| Task Information Page 5
| Library . . : QPEXDATA
| Member. . . : STATSF
| Description : BLANK
| Task ID Job/Task Name Pool Priority Existence Elapsed Time (us) CPU (us) CPU
| Start/End Percent
| ----------------------------------------------------------------------------------------------------------------------------------
| 6739 QPADEVðð29 FOLEY 1ð1ð29 4 163 Y Y 41897112 822ð96 2ð.97
| 1 CFINTð1 ð ð Y Y 4257ððð8 3ð98472 79.ð3
| ----------------------------------------------------------------------------------------------------------------------------------

| Figure 11-4. Task Information Report

The Task Information Report shows the following values:

Library The library the database file is in.
Member The database file member containing the data.
Description The description of the data that was collected.
Task ID The system identifier for the task
Job/Task Name The name of the task or job under which the data was run.
Pool The system pool that the job or task in on the system.
Priority The relative LIC priority that the job or task runs at on the system.
Note: The priority column here is not the same as the job priority that
is shown on the Work with Active Jobs display. You can subtract
140 from the non-SLIC tasks to find the RUNPTY value used on
AS/400.
Existence Start Indicates if the job or task existed at the start of the collection
(Y/N)
Existence End Indicates if the job or task existed at the end of the collection (Y/N)
Elapsed Time (us) The elapsed time (in microseconds) that the job or task existed
during the collection
CPU (us) The total amount of CPU time used (in microseconds) by the job or task
during the collection.
CPU Percent The percentage of CPU used by this job or task as compared to the
total CPU used by all the jobs or tasks in the collection.

| The Priority values that are shown do not correspond to the Run priority
| (RUNPTY) parameter value. However, for OS/400 jobs with priority values of 1-99,
| you can subtract 140 to correspond to the RUNPTY value.

| RUNPTY(15) is a typical priority for spooled writer jobs (class QSPL).

| For Licensed Internal Code (LIC) tasks, the user cannot change them. In most
| cases LIC task priorities are higher than OS/400 jobs. However, some LIC tasks
| run at the same priority as the user job for which they are performing a function.

| The disk drive tasks that start with prefix DBI or DBL typically run under the
| RUNPTY value of the OS/400 job for which they are performing the function.

11-14 Performance Tools V4R2

 Report-Specific Sections
This section shows examples of the main reports and also report sections that are
specific to certain reports.

Summary Information
Summary information provides a subset of the information shown in the main
reports. The Profile Report and the Statistics Report have their own Summary Infor-
mation. The Trace Report does not include a Summary Information Report.

Performance Explorer Report 1/21/97 13:39:1ð

Profile CPU Summary Information Page
3
Library . . . . : COOL
Member. . . . . : RBPROF2PGM
Description . . : RBPROF-CMDCSTPEXH (CLCSTPEXHI, CSTPEX)
Total CPU . . . : 848ð864
Job CPU . . . . : 8256856 97.4 %
Task CPU. . . . : 224ðð8 2.6 %
-----------------------
Total Samples . : 7664
Total Hits. . . : 11ð8 14.5 %

Figure 11-5. Example for *PROFILE CPU Summary Information

The Profile CPU Summary Information Report shows the following values:
Library The library the database file is in.
Member The database file member containing the data.
Description The description the user associated with the collected data
Total CPU The total amount of CPU used by the tasks and jobs that were collected
on (not the whole system CPU).
Job CPU The total amount of CPU used by the jobs that were collected on.
Task CPU The total amount of CPU used by the tasks that were collected on.
Total Samples The total number of samples collected during a session.
Total Hits The total number of samples that occurred within the programs the user
specified.

Performance Explorer Report 11/ð7/97 9:54:47

Stats CPU Summary Information Page 3
Library . . . . . : QPEXDATA
Member. . . . . . : STATSF
Job name. . . . . : ALL JOBS/TASKS IN SESSION
Description . . . : BLANK
Total Raw CPU . . : 392ð568
Overhead Removed. : 112381
Total CPU . . . . : 38ð8187
Task CPU. . . . . : 3ð98472 81.4 %
Job CPU . . . . . : 7ð9715 18.6 %
-----------------------
Pgm/Mod CPU. . : 31ð419 8.2 %
Unknown CPU. . : 399296 1ð.5 %

Figure 11-6. *STATS CPU Summary Information

The Statistics Summary shows the same fields as the Profile Summary with the
addition of the following values:
| Job name The job name, user ID, and job number. ALL JOBS/TASKS IN
| SESSION is a special value.

Chapter 11. Performance Explorer 11-15

 Total Raw CPU The total amount of CPU used by tasks and jobs that was col-
lected on (including any collector overhead).
Overhead Removed The difference between the total raw CPU and total CPU (in
other words, adjusted CPU).
Total CPU The total amount of CPU used by tasks and jobs that were collected on
(less any collection overhead that could be removed).
Pgm/Mod CPU The total CPU of the programs and modules that were collected on
and reported on in the STATS INFORMATION section of the report.
Unknown CPU The difference between the Job CPU and the Pgm/Mod CPU.

Library Information
Library information - shows collection information for each library. Also provides
data on call and complex MI counts, CPU utilization, and disk I/O operations. This
section is available for the statistical report only.

The library section identifies the libraries that contained the programs or modules
that were active during the collection period. All CPU usage and disk I/O operation
statistics for all the programs or modules in a specific library are totaled for that
library. It is common to have a cumulative CPU percent total that is higher than
99.9%. In those cases, you see a CPU percent value of ****. The **** value is con-
sidered normal in most cases. Figure 11-7 on page 11-17 shows a Library Section
that summarizes the CPU and disk I/O activity at the library level.

| One area of interest is the **Unknown category. You can start and stop a collection
| at any time. You receive resource usage data, but depending on what the programs
| in a job are doing, you may not see it accounted for accurately. When this
| happens, the results end up in the **Unknown category and not in the program that
| you thought was using all the resources.

| The shorter the time period that you run a collection, the greater the percentage
| allocated to **Unknown may be. This occurs because performance explorer collects
| data from the entry to a program and from the exit from a program. If the program
| is already entered when you start the collection, the data that is collected is not
| allocated to that specific program. Instead, the data gets counted and put into a
| counter called **Unknown.

The Statistics Report, at the library level, shows the following values:
Library The library the database file is in.
Member The database file member containing the data.
Description The description of the data that was saved.
Name The Name of the library for which the statistics are being shown.
Times Called The number of times programs in that library were called.
Calls Made The number of calls programs in the library made
MI CPLX Issued The number of MI complex instructions that were called by a
program or procedure. MI complex instructions are the architected MI
instructions of AS/400. They are identified in the report with a single “*”
in front of the instruction name.
Inline Stats The statistics that were incurred directly by programs in the library.

11-16 Performance Tools V4R2

 Performance Explorer Report 11/ð7/97 9:54:47
Statistics Information Page 3
Library Section
Library . . : QPEXDATA
Member. . . : STATSF
Job name. . : ALL JOBS/TASKS IN SESSION
Description : BLANK
+----------------- Inline Stats ------------++-------------- Cumulative Stats -----------+
Times Calls MI CPLX CPU DB DB NDB NDB CPU DB DB NDB NDB Call
Name Called Made Issued (us) / % SIO AIO SIO AIO (us) / % SIO AIO SIO AIO Level
----------------------------------------------------------------------------------------------------------------------------------
\\LIC Task ð ð ð 3,ð98,472 81.4 ð ð ð ð 3,ð98,472 81.4 ð ð ð ð ð
\\Unknown ð ð ð 399,295 1ð.5 ð ð ð ð 399,295 1ð.5 ð ð ð ð ð

Figure 11-7. *STATS Information Library Section

CPU (us) The amount of CPU in microseconds used by programs in

the library
% The percentage of CPU used as compared to the Total CPU
found in the summary section.
DB SIO The number of database synchronous I/O operations per-
formed by programs in the library.
DB AIO The number of database asynchronous I/O operations per-
formed by programs in the library.
NDB SIO The number of non-database synchronous I/O operations
performed by programs in the library.
NDB AIO The number of non-database asynchronous I/O operations
performed by programs in the library.
| Cumulative Stats The statistics that were incurred both directly and indirectly by
| programs in the library. This can occur through calls to other programs
| in the other libraries in this list.
| Note: At the library level, cumulative statistics can be greater than
| 100%. In this example, the \\\\ means that the percent is
| greater than 100%.
| CPU (us) The cumulative amount of CPU in microseconds that is used
| by programs in the library and other programs they called.
| % The percentage of cumulative CPU that is used as compared
| to the Total CPU that is found in the summary section.
| DB SIO The cumulative number of database synchronous I/O oper-
| ations that are performed by programs in the library and
| other programs they called.
| DB AIO The cumulative number of database asynchronous I/O oper-
| ations performed by programs in the library and other pro-
| grams they called.
| NDB SIO The cumulative number of non-database synchronous I/O
| operations that are performed by programs in the library and
| other programs they called.
| NDB AIO The cumulative number of non-database asynchronous I/O
| operations performed by programs in the library and other
| programs they called.

Chapter 11. Performance Explorer 11-17

| Call Level Shows the invocation call level in a hierarchical statistics col-
| lection. Specify DATAORG(*HIER) on the ADDPEXDFN
| command to show the data in a hierarchical format.

| Figure 11-8 shows a sample Library section from the Profile Information Report.

| Performance Explorer Report 6/ð1/95 1ð:34:15

| Profile Information Page 3

| Library Section
| Library . . : COOK
| Member. . . : RBPROF2PGM
| Description : RBPROF-CMDCSTPEXH (CLCSTPEXHI, CSTPEX)
| Histogram Hit Hit Cum Start Map Stmt Name
| Cnt % % Addr Flag Numb
| ----------------------------------------------------------------------------------------------------------------------------------
| \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ 11ð8 1ðð.ð 1ðð.ð 22B55B7DFDðð236ð MP 7 PFREXP

| Figure 11-8. Example for *PROFILE Library Section

The Profile Information Report shows the following values:

Library The library the database file is in.
Member The database file member containing the data.
Description The description of the data that was saved.
Histogram The histogram graphically illustrates the number of sample hits against
this entry as compared with the total number of sample hits.
Hit Cnt The number of samples that occurred within this entry
Hit % The percent of hits for this line as compared to the total number of
sample hits.
Cum % The cumulative Hit % of this entry and all preceding entries.
| Start Addr The address of the first instruction. The Start Addr column is only filled
| in for the Profile Information Report.
Map Flag The map flag indicates what this entry corresponds to. The possible
values for the map flag are:
Ÿ DS (distributed statement). This line represents multiple source
statements that were optimized into a single instruction.
Ÿ SS (single statement). This line represents a single source state-
ment.
Ÿ MS (multiple statement). This line represents multiple source state-
ments.
Ÿ MP (multiple procedures). This line represents multiple procedures.
Stmt Numb The MI statement number of the first instruction of this entry.
Notes:
1. To get a detailed Profile Information Report that shows the utilization
of your HLL program statements, you should specify
TYPE(*PROFILE) on the PRTPEXRPT command and summarize
the collected data by PROFILEOPT(*BLANK) or
PROFILEOPT(*STATEMENT).

11-18 Performance Tools V4R2

 2. For information on how the source code maps to the statement
number, see “Mapping OPM High-Level Language (HLL) Statements
to Source Code” on page 11-35.
Name The name of the program, module, and procedure associated with this
entry. If the report is summarized at the Module level, then the proce-
dure will be left off. If the report is summarized at the Program level,
then the module and the procedure names will be left off.

Statistics Report
The Statistics Report provides general performance statistics to help identify
problem areas. In partaicular, the statistics details show potential program perform-
ance trouble spots of a job or system.

The report also shows a variety of other use statistics, such as invocations and
number of disk I/Os. From the Statistics Report, you should be able to determine
how much resource the programs in your application are using. Using this informa-
tion, you can determine if there is a performance problem that requires further
investigation.

| If you notice a single library with a high level of CPU utilization or DASD I/O oper-
| ations, you might want to focus on programs in that library. Libraries that you might
| see could include the following:
| Ÿ QSYS, which stores most of the OS/400 modules.
| Ÿ QTCP, which provides TCP/IP support.
| Ÿ QIJS, which provides the Job Scheduler for OS/400 support.
| Ÿ QBRM, which provides the Backup Recovery and Media Services support.

See the “Library Information” on page 11-16 section for a discussion of the column
descriptions.

Chapter 11. Performance Explorer 11-19

 Performance Explorer Report
Statistics Information Page 6
Library . . : QPEXDATA
Member. . . : STATSF
Job name. . : ALL JOBS/TASKS IN SESSION
Description : BLANK
+----------------- Inline Stats ------------++-------------- Cumulative Stats -----------+
Times Calls MI CPLX CPU DB DB NDB NDB CPU DB DB NDB NDB Call
Name Called Made Issued (us) / % SIO AIO SIO AIO (us) / % SIO AIO SIO AIO Level
----------------------------------------------------------------------------------------------------------------------------------
\\LIC Task ð ð ð 3,ð98,472 81.4 ð ð ð ð 3,ð98,472 81.4 ð ð ð ð ð
\\Unknown ð ð ð 399,295 1ð.5 ð ð ð ð 399,295 1ð.5 ð ð ð ð ð
\CRTS 9 ð ð 23,365 ð.6 ð ð 14 ð 23,365 ð.6 ð ð 14 ð ð
\DEQWAIT 1ð ð ð 22,5ð5 ð.6 ð ð 2 ð 22,5ð5 ð.6 ð ð 2 ð ð
\DESS 9 ð ð 13,7ð1 ð.4 ð ð 6 ð 13,7ð1 ð.4 ð ð 6 ð ð
\RSLVSP 37 ð ð 11,174 ð.3 ð ð ð ð 11,174 ð.3 ð ð ð ð ð
\MATPRMSG 68 ð ð 9,471 ð.2 ð ð ð ð 9,471 ð.2 ð ð ð ð ð
QWSPUT 23 6 26 9,157 ð.2 ð ð ð ð 15,3ð5 ð.4 ð ð ð ð ð
QSFPUT 15 1 ð 8,151 ð.2 ð ð ð ð 17,11ð ð.4 ð ð 3 ð ð
QMHRCVPM 24 3 115 7,611 ð.2 ð ð ð ð 71,841 1.9 ð ð 3 ð ð
QPTPRCSS 1 57 ð 7,591 ð.2 ð ð ð ð 66,ð43 1.7 ð ð 4 ð ð
\SETACST 2ð ð ð 7,517 ð.2 ð ð 6 2 7,517 ð.2 ð ð 6 2 ð
QYPEENDP 3 2ð 2ð 7,428 ð.2 ð ð 3 ð 142,7ð8 3.7 ð 2 4ð 4 ð
QMHGSD 6 48 24 7,192 ð.2 ð ð ð ð 6ð,865 1.6 ð ð 3 ð ð
\FNDINXEN 26 ð ð 7,174 ð.2 ð ð 1 ð 7,174 ð.2 ð ð 1 ð ð
QWSGET 17 9 17 7,ð99 ð.2 ð ð ð ð 34,884 ð.9 ð ð 2 ð ð
\SNDPRMSG 21 ð ð 6,886 ð.2 ð ð ð ð 6,886 ð.2 ð ð ð ð ð
QUIINMGR 6 12 ð 5,717 ð.2 ð ð ð ð 29,ð2ð ð.8 ð ð 2 ð ð
\MATPTR 1ð1 ð ð 5,462 ð.1 ð ð ð ð 5,462 ð.1 ð ð ð ð ð
\MODPRMSG 46 ð ð 5,251 ð.1 ð ð ð ð 5,251 ð.1 ð ð ð ð ð
QCAFLD 17 ð ð 4,976 ð.1 ð ð ð ð 4,976 ð.1 ð ð ð ð ð
QUIMGFLW 5 33 16 4,928 ð.1 ð ð ð ð 71,679 1.9 ð ð 7 ð ð
QLIDLOBJ 2 4 16 4,317 ð.1 ð ð 5 ð 17,ð39 ð.4 ð ð 13 ð ð
\MODS1 1ð ð ð 4,127 ð.1 ð ð ð ð 4,127 ð.1 ð ð ð ð ð
QUIEXFMT 6 23 3 4,ð44 ð.1 ð ð ð ð 5ð,525 1.3 ð ð 3 ð ð
\MATINVIF 135 ð ð 4,ðð2 ð.1 ð ð ð ð 4,ðð2 ð.1 ð ð ð ð ð
QUILIST 26 ð 8 3,885 ð.1 ð ð ð ð 4,449 ð.1 ð ð ð ð ð
QCMDEXC 4 22 18 3,571 ð.1 ð ð ð ð 3ð,858 ð.8 ð ð 3 ð ð
QCADRV2 11 4ð 1 3,439 ð.1 ð ð ð ð 16,1ð6 ð.4 ð ð ð ð ð
QMHSNDPM 1ð ð 43 3,239 ð.1 ð ð 1 ð 9,7ð1 ð.3 ð ð 1 ð ð
\MATSOBJ 2ð ð ð 3,ð66 ð.1 ð ð 1 ð 3,ð66 ð.1 ð ð 1 ð ð
QUIVPMGR 27 ð ð 2,933 ð.1 ð ð ð ð 2,933 ð.1 ð ð ð ð ð
QLICKOBJ 4 6 6 2,9ð7 ð.1 ð ð ð ð 11,398 ð.3 ð ð 2 ð ð
QCAPOS 7 ð ð 2,879 ð.1 ð ð ð ð 2,879 ð.1 ð ð ð ð ð
\MODADR 2 ð ð 2,838 ð.1 ð ð ð 2 2,838 ð.1 ð ð ð 2 ð
QWSSFLCT 5 ð ð 2,829 ð.1 ð ð ð ð 2,829 ð.1 ð ð ð ð ð
\REQIO 1ð ð ð 2,753 ð.1 ð ð ð ð 2,753 ð.1 ð ð ð ð ð
QUIOCNV 6 ð ð 2,736 ð.1 ð ð ð ð 2,736 ð.1 ð ð ð ð ð
QLIINSRT 2 1ð 28 2,645 ð.1 ð ð ð ð 27,4ð9 ð.7 ð 2 3 4 ð
\CRTDOBJ 1 ð ð 2,6ð5 ð.1 ð ð 1 ð 2,6ð5 ð.1 ð ð 1 ð ð
\LOCK 2ð ð ð 2,525 ð.1 ð ð ð ð 2,525 ð.1 ð ð ð ð ð
QMHRMVPM 14 ð 64 2,495 ð.1 ð ð ð ð 6,229 ð.2 ð ð ð ð ð
QSZGTPRD 4 6 18 2,484 ð.1 ð ð ð ð 8,3ðð ð.2 ð ð ð ð ð
QPTGTINP 3 12 6 2,226 ð.1 ð ð 1 ð 8,553 ð.2 ð ð 1 ð ð
QT3REQIO 13 ð 3ð 2,ð35 ð.1 ð ð ð ð 27,552 ð.7 ð ð 2 ð ð

Figure 11-9. *STATS Information

Profile Report
The Profile Report provides output to show relative CPU time by procedure. This
allows the user to identify where to focus efforts to provide overall performance of
the application or program.
Note: You can also summarize profile counts at a procedure, module, or program
level.

See the “Library Information” on page 11-16 section for a discussion of the column
descriptions.

11-20 Performance Tools V4R2

| In this example, you can see that three statements were responsible for 71.75% of
| the CPU costs. It would not be worthwhile to investigate other statements that
| reported smaller percentages. In a real application program, the profile information
| can show a high percentage of CPU cost for a single statement or similar percent-
| ages for more than one statement. Look at the program source code to determine if
| the high CPU cost is due to a single statement or a group of statements. A group
| of statements can indicate a processing loop.

| Performance Explorer Report 1/21/97 13:39:1ð

| Profile Information Page 12
| Library . . : COOK
| Member. . . : RBPROF2PGM
| Description : RBPROF-CMDCSTPEXH (CLCSTPEXHI, CSTPEX)
| Histogram Hit Hit Cum Start Map Stmt Name
| Cnt % % Addr Flag Numb
| ----------------------------------------------------------------------------------------------------------------------------------
| \\\\\\\\\\\ 349 31.5 31.5 22B55B7DFDðð2E48 MP 45 CSTPEX
| \\\\\\\ 243 21.9 53.4 22B55B7DFDðð2DAð MP 42 CSTPEX
| \\\\\\ 2ð2 18.2 71.7 22B55B7DFDðð3E4ð MP 1ð6 CSTPEX
| \\ 82 7.4 79.1 22B55B7DFDðð2F2ð MP 46 CSTPEX
| \\ 81 7.3 86.4 22B55B7DFDðð2D6C MP 41 CSTPEX
| \ 53 4.8 91.2 22B55B7DFDðð3ECC MP 1ð7 CSTPEX
| \ 45 4.1 95.2 22B55B7DFDðð2E2C MP 44 CSTPEX
| 15 1.4 96.6 22B55B7DFDðð2F9ð MP 47 CSTPEX
| 1ð ð.9 97.5 22B55B7DFDðð3E38 MP 1ð5 CSTPEX
| 7 ð.6 98.1 22B55B7DFDðð2FAð MP 5ð CSTPEX
| 4 ð.4 98.5 22B55B7DFDðð236ð MP 7 CSTPEX
| 4 ð.4 98.8 22B55B7DFDðð25ðð MP 14 CSTPEX
| 2 ð.2 99.ð 22B55B7DFDðð23F4 MP 1ð CSTPEX
| 2 ð.2 99.2 22B55B7DFDðð2D48 MP 4ð CSTPEX
| 2 ð.2 99.4 22B55B7DFDðð31ðð MP 55 CSTPEX
| 2 ð.2 99.5 22B55B7DFDðð3288 MP 69 CSTPEX
| 2 ð.2 99.7 22B55B7DFDðð454C MP 128 CSTPEX
| 1 ð.1 99.8 3567AA4ð1Dðð24B8 MP 54ðð CLCSTPEXHI
| 1 ð.1 99.9 22B55B7DFDðð24Bð MP 13 CSTPEX
| 1 ð.1 1ðð.ð 22B55B7DFDðð2CCð MP 38 CSTPEX

| Figure 11-10. Example for *PROFILE Information

Trace Report
The Trace Report provides a historical trace of performance activity generated by
one or more jobs or tasks on the system.

Figure 11-11 on page 11-22 shows a sample Trace report.

Chapter 11. Performance Explorer 11-21

 572848
Performance Explorer Report
Trace Information Page 4
Library . . : QPEXDATA
Member. . . : SFCALLRTN
Description : BLANK
Time Stamp. : 15.ð7.56.ð83ððð Task ID: ððððð9F5 Name: QPADEVððð4 FOLEY ð54858 Run Time (us): 484192 Percent: 84.52
P = Processor Number M = Missed Event Indicator Obj = Object Seg = Segment
T ST = Type Subtype NAGP = Non Access Group Pages NPgs = Number of Pages Unit = DASD Unit/Sub Unit
Sector = DASD Sector PI = Pool ID AI = ASP ID SKP = DASD Skip Operation
Span = DASD Sectors Spanned EXID = Exception ID IEID = IMPI Exception ID XCH = DASD Exchange
PTY = Apparent Task Priority WODSC = Wait Obj Description RS = Wait Obj Reason
PREFIX = S: Stealable Page PREFIX = A: PAG Data PREFIX = D: Data Base Data PREFIX = M: Mirrored DASD
PREFIX = P: Permanent Segment PREFIX = T: Temporary Segment PREFIX = E: E=R Address
Address Offset Object Name Obj Seg PRE NPgs LIC-Pgm--Offset MI-Pgm-----Offset NAGP PI AI
T ST T ST FIX Unit Sector Span SKP XCH EXID IEID
ss.mmm P M Task ID Parent-Pgm HLL-No CurrentPgm RC Delta Run Cycles Event PTY WaitSleep Cycles WODSC RS SNDTSK
----------------------------------------------------------------------------------------------------------------------------------
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð15F \RSLVSP ð ð MISTR RESOLVE SYSTEM POINTER
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð16ð \RSLVSP ð ð MIEND RESOLVE SYSTEM POINTER
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð164 \RSLVSP ð ð MISTR RESOLVE SYSTEM POINTER
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð165 \RSLVSP ð ð MIEND RESOLVE SYSTEM POINTER
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð167 \TESTAU ð ð MISTR TEST AUTHORITY
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð167 \TESTAU ð ð MIEND TEST AUTHORITY
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð17E \FNDINXEN ð ð MISTR FIND INDEPENDENT INDEX ENTRY
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð17F \FNDINXEN ð ð MIEND FIND INDEPENDENT INDEX ENTRY
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð491 \MATINVIF ð ð MISTR MATERIALIZE INVOCATION INFO.
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð492 \MATINVIF ð ð MIEND MATERIALIZE INVOCATION INFO.
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð2ðF \MATINVIF ð ð MISTR MATERIALIZE INVOCATION INFO.
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð21ð \MATINVIF ð ð MIEND MATERIALIZE INVOCATION INFO.
56.ð83 7 ð ððððð9F5 QMHSNDPM ððð6AD \SNDPRMSG ð ð MISTR SEND PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHSNDPM ððð6AE \SNDPRMSG ð ð MIEND SEND PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHSNDPM ð ð EXIT
56.ð84 7 ð ððððð9F5 QYPESTRP ð ð EXIT QYPESTRP/_CXX_PEP
56.ð84 7 ð ððððð9F5 QCMD ððð182 \TESTEXCP ð ð MISTR TEST EXCEPTION
56.ð84 7 ð ððððð9F5 QCMD ððð182 \TESTEXCP ð ð MIEND TEST EXCEPTION
56.ð84 7 ð ððððð9F5 QMHRCVPM ð ð ENTRY
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð4CC \MATPRMSG ð ð MISTR MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð4CD \MATPRMSG ð ð MIEND MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð4E3 \MATINVIF ð ð MISTR MATERIALIZE INVOCATION INFO.
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð4E4 \MATINVIF ð ð MIEND MATERIALIZE INVOCATION INFO.
56.ð84 7 ð ððððð9F5 QMHRCVPM ðððð36 \MATPRMSG ð ð MISTR MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHRCVPM ðððð37 \MATPRMSG ð ð MIEND MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð44C \MATINVIF ð ð MISTR MATERIALIZE INVOCATION INFO.
56.ð84 7 ð ððððð9F5 QMHRCVPM ððð44D \MATINVIF ð ð MIEND MATERIALIZE INVOCATION INFO.
56.ð84 7 ð ððððð9F5 QMHGSD ð ð ENTRY
56.ð84 7 ð ððððð9F5 QMHGSD ððð827 \MATPRMSG ð ð MISTR MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHGSD ððð828 \MATPRMSG ð ð MIEND MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QUIVPMGR ð ð ENTRY
56.ð84 7 ð ððððð9F5 QUIVPMGR ð ð EXIT
56.ð84 7 ð ððððð9F5 QUIVPMGR ð ð ENTRY
56.ð84 7 ð ððððð9F5 QUIVPMGR ð ð EXIT
56.ð84 7 ð ððððð9F5 QUILIST ð ð ENTRY
56.ð84 7 ð ððððð9F5 QUILIST ð ð EXIT
56.ð84 7 ð ððððð9F5 QMHGSD ððð819 \MATPRMSG ð ð MISTR MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHGSD ððð81A \MATPRMSG ð ð MIEND MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHGSD ððð819 \MATPRMSG ð ð MISTR MATERIALIZE PROCESS MESSAGE
56.ð84 7 ð ððððð9F5 QMHGSD ððð81A \MATPRMSG ð ð MIEND MATERIALIZE PROCESS MESSAGE

Figure 11-11. Trace Information Report

The Trace Report shows the following values:

Library The library the database file is in.
Member The database file member containing the data.
Description The description of the data that was saved.
Time stamp The full time of day delta from the start of the session for the first
event on each page of the report.

11-22 Performance Tools V4R2

Task ID The task ID (in hexadecimal) of the events on each page.
Note: This information appears only when the sort by *TASK value is
specified on the PRTPEXRPT command.
Name The name of the task associated with the events on each page.
Note: This information appears only when the sort by *TASK value is
specified on the PRTPEXRPT command.
Run Time (us) The total run time in microseconds of the task associated with the
events on each page.
Note: This information appears only when the sort by *TASK value is
specified on the PRTPEXRPT command.
Percent The percent value of the total run time of the task associated with the
events on each page.
Note: This information appears only when the sort by *TASK value is
specified on the PRTPEXRPT command.
ss.mmm The number of seconds.milliseconds of the event from the start of the
session.
P (Processor Number) The number of the processor.
M Missed event indicator. Events are missing because of the following:

B The collection mechanism overloaded.

U The collection mechanism is not available.
* Unknown reasons.
Task ID The task ID (in hexadecimal) of the task responsible for the event.
Program The name of the program associated with the event.
HLL-No The high-level language statement number (in hexadecimal) of the
program where the MI complex instruction was issued.
CurrentPgm The name of the MI complex instruction or program associated with
the event.
RC Delta The difference between the current event run cycle value and the pre-
vious event run cycle value.
Run Cycles The run cycle value for the event. Run cycles are the number of non-
idle CPU cycles and are accumulated on a per job or task basis.
Event The 5-character event abbreviation that identifies what event occurred.

In addition to the previous columns that are shown for the Trace Report, you see
the following columns when you specify the TRCTYPE parameter.
Address Offset The hexadecimal representation of an address associated with the
event.
Object Name The name of the object associated with the event.
Obj T ST The object type and subtype (in hexadecimal) of the object associated
with the event.

Chapter 11. Performance Explorer 11-23

 Seg T ST The segment type and subtype (in hexadecimal) of the address associ-
ated with the event.
PREFIX Character flags that give more detail of the object associated with the
event.
S Stealable page
A PAG data
D Database data
M Mirrored DASD
P Permanent segment
T Temporary segment
E E=R address
NPgs The number of pages requested by the event.
LIC-Pgm--Offset The Licensed Internal Code (LIC) program identifier and instruc-
tion offset associated with the event.
MI-Pgm Offset The program name and instruction offset associated with the event.
NAGP The number of non-activation group pages requested by the event.
PI The pool identifier associated with the event.
AI The auxiliary storage pool identifier associated with the event.
Unit The DASD unit number subunit number (in hexadecimal).
Sector The DASD sector address associated with the event.
Span The span of the DASD request associated with the event.
SKP XCH Y/N columns indicating whether the DASD event was a skip operation or
an exchange operation.
EXID Exception identifier column (in hexadecimal).
IEID IMPI Exception identifier column (in hexadecimal).

| Trace Event Descriptions: The following tables describe each event available
| when you specify SLTEVT(*YES) on the ADDPEXDFN command. The tables also
| indicate the relationships between the TRCTYPE parameter and the events that are
| included in the performance explorer definition.

| The data that is collected for the BASE events are placed in the QAYPEDTIDX and
| QAYPEBASE files.

| Table 11-2 (Page 1 of 3). BASE event descriptions

| Event TRCTYPE Description
| PRCCRT Process Create. An event is generated whenever an
| AS/400 process is created during an active perform-
| ance explorer collection session.
| PRCDLT Process Delete. An event is generated whenever an
| AS/400 process ends during an active performance
| explorer collection session.

11-24 Performance Tools V4R2

| Table 11-2 (Page 2 of 3). BASE event descriptions
| Event TRCTYPE Description
| TASKCRT Task Create. An event is generated whenever an
| AS/400 SLIC task is created during an active perform-
| ance explorer collection session.
| TASKDLT Task Delete. An event is generated whenever an
| AS/400 SLIC task ends during an active performance
| explorer collection session.
| TASKSWTIN *TASKSWT Task Switch In. An event is generated whenever an
| AS/400 task is switched in and given processor time
| during an active performance explorer collection
| session.
| TASKSWTOUT *TASKSWT Task Switch Out. An event is generated whenever an
| AS/400 task is switched out from the processor, due to
| time-slice end, during an active performance explorer
| collection session.
| PMCO ALL TRCTYPES Performance Measurement Counter Overflow. An
| event is generated whenever this hardware counter
| overflows during an active performance explorer col-
| lection session. This hardware counter is used in con-
| junction with the Sample rate parameter and can be
| programmed to overflow and generate an event at the
| specified sample rate interval.
| TASKSWTOUTQ *TASKSWT Task Switch Out Queueing. An event is generated
| whenever an AS/400 task is switched out from the
| processor, due to a lock condition, during an active
| performance explorer collection session.
| TASKSWTOUTINT *TASKSWT Task Switch Out Interrup. An event is generated when-
| ever an AS/400 task is switched out from the
| processor, due to an interrupt, during an active per-
| formance explorer collection session.
| TASKAVAIL *TASKSWT Task Available for Dispatch. An event is generated
| whenever an AS/400 task is available to be dispatch to
| the processor during an active performance explorer
| collection session.
| ACTGRPACTPGM *BASIC, *PGMACT Activation Group Activate Program. An event is gener-
| ated whenever an ILE program activation begins
| during an active performance explorer collection
| session.
| ACTGRPCRT *BASIC, *PGMACT Activation Group Create. An event is generated when-
| ever an ILE program activation group is created during
| an active performance explorer collection session.
| ACTGRPDLT *BASIC, *PGMACT Activation Group Delete. An event is generated when-
| ever an ILE program activation group is destroyed
| during an active performance explorer collection
| session.
| EXCP Exception. An event is generated whenever an excep-
| tion occurs during an active performance explorer col-
| lection session.
| MIEXCP Mi Exception. An event is generated whenever an MI
| exception occurs during an active performance
| explorer collection session.

Chapter 11. Performance Explorer 11-25

| Table 11-2 (Page 3 of 3). BASE event descriptions
| Event TRCTYPE Description
| EXTINTSTR External Interrupt Start. An event is generated when-
| ever a start of an external interrrupt occurs during an
| active performance explorer collection session.
| EXTINTEND External Interrupt End. An event is generated when-
| ever an end of an external interrupt occurs during an
| active performance explorer collection session.
| DCRINTSTR Decrement Interrupt Star. An event is generated when-
| ever a start of a decrement interrupt occurs during an
| active performance explorer collection session.
| DCRINTEND Decrement Interrupt End. An event is generated when-
| ever an end of a decrement interrupt occurs during an
| active performance explorer collection session.
| PGMDBG *BASIC Program Debug. This event is activated through the
| Dedicated Service Tools (DST) interface

| Auxilary storage management (ASM) is responsible for the management of the

| segment spaces for the single-level storage model of the AS/400.

| The data that is collected for the ASM events are placed in the QAYPETIDX and
| QAYPEASM files.

| Table 11-3. Auxiliary storage management event descriptions

| Event TRCTYPE Description
| CRTSEG *BASIC, *DSKSTG, *VRTADR Create Segment. An event is generated when-
| ever a segment is created during an active per-
| formance explorer collection session.
| DLTSEG *BASIC, *DSKSTG Destroy Segment. An event is generated when-
| ever a segment is destroyed during an active
| performance explorer collection session.
| EXDSEG *BASIC, *DSKSTG Extend Segment. An event is generated when-
| ever a segment is extended during an active
| performance explorer collection session.
| FNDSEGSIZ *BASIC, *DSKSTG Find Segment Size. An event is generated
| whenever a request to find the segment size
| occurs during an active performance explorer
| collection session.
| TRUNCSEG *BASIC, *DSKSTG Truncate Segment. An event is generated
| whenver a segment is truncated during an active
| performance explorer collection session.

| Disk events trace I/O activity with the disk storages devices of the AS/400.

| The data that is collected for the Disk events are placed in the QAYPETIDX and
| QAYPEDASD files.

11-26 Performance Tools V4R2

| Table 11-4. Disk event descriptions
| Event TRCTYPE Description
| READSTR *BASIC, *DSKIO1, *DSKIO2, Read Start. An event is generated whenever a
| *TASKSWT start of a physical read occurs during an active
| performance explorer collection session.
| READEND *DSKIO1, *DSKIO2, *TASKSWT Read End. An event is generated whenever a
| complete of a physical read occurs during an
| active performance explorer collection session.
| WRTSTR *BASIC, *DSKIO1, *DSKIO2, Write Start. An event is generated whenever a
| *TASKSWT start of a physical write occurs during an active
| performance explorer collection session.
| WRTEND *DSKIO1, *DSKIO2, *TASKSWT Write End. An event is generated whenever a
| complete of a physical write occurs during an
| active performance explorer collection session.
| PGREADSTR Page Read Start. An event is generated when-
| ever a start of a page read occurs during an
| active performance explorer collection session.
| PGREADEND Page Read End. An event is generated when-
| ever a complete of a page read occurs during
| an active performance explorer collection
| session.
| PGWRTSTR Page Write Start. An event is generated when-
| ever a start of page write occurs during an
| active performance explorer collection session.
| PGWRTEND Page Write End. An event is generated when-
| ever a complete of a page write occurs during
| an active performance explorer collection
| session.

| A page fault occurs whenever an address is referred to and is not in the main
| storage of the AS/400 system.

| The data that is collected for the page fault events are placed in the QAYPETIDX
| and QAYPEPGFLT files.

| Table 11-5. Page fault event descriptions

| Event TRCTYPE Description
| STR *BASIC, *DSKIO2 Start of Fault. An event is generated whenever a start
| of a page fault occurs during an active performance
| explorer collection session.
| ENDOK End of Fault: OK. An event is generated whenever a
| successful completion of a page fault occurs during an
| active performance explorer collection session.
| ENDERR End of Fault: Error. An event is generated whenever
| an unsuccesful completion of a page fault occurs
| during an active performance explorer collection
| session.

| The mangagement of the MI interrupts are for the jobs or processes running on the
| AS/400 system

Chapter 11. Performance Explorer 11-27

| The data that is collected for the RMPM events are placed in the QAYPETIDX and
| QAYPERMPM files.

| Table 11-6. Job or Process Events

| Event TRCTYPE Description
| LWSTR Machine Interface Long Wait Start. An event is gener-
| ated whenever a start of a long wait occurs during an
| active performance explorer collection session.
| LWEND Machine Interface Long Wait End. An event is gener-
| ated whenever a completion of a long wait occurs
| during an active performance explorer collection
| session.
| INELIGIBLE Machine Interface Process Ineligible. An event is gen-
| erated whenever a process moves from active state to
| ineligible during an active performance explorer col-
| lection session.
| ACTIVE Machine Interface Process Active. An event is gener-
| ated whenever a process moves from ineligible or wait
| to active during an active performance explorer col-
| lection session.
| INTERRUPT Machine Interface Process Interrupt. An event is gen-
| erated whenever a process interrupt occurs during an
| active performance explorer collection session.
| MPLPOOLCHG Machine Interface MPL Pool Change. An event is gen-
| erated whenever the MPL pool changes during an
| active performance explorer collection session.
| TOBCHMPLPOOL Machine Interface Process To Batch MPL Pool. An
| event is generated whenever an MI process is being
| implicitly moved to the batch MPL pool defined for it
| during an active performance explorer collection
| session.
| TSLEND Machine Interface Timeslice End. An event is gener-
| ated whenever a process timeslice end occurs during
| an active performance explorer collection session.

| The mangagement of the seizes and locks used for resource serialization that
| occurr on the AS/400 system.

| The data that is collected for the RMSL events are placed in the QAYPETIDX and
| QAYPERMSL files.

| Table 11-7. Seize Lock Events

| Event TRCTYPE Description
| LWEND Sieze Lock Long Seize Wait End. An event is gener-
| ated whenever a long seize wait completes during an
| active performance explorer collection session.

| The SAR events (give description here)

| The data that is collected for the SAR events are placed in the QAYPETIDX and
| QAYPESAR files.

11-28 Performance Tools V4R2

| Table 11-8 (Page 1 of 2). Segment Address Register Events
| Event TRCTYPE Description
| CLR *BASIC Clear. An event is generated whenever a storage
| clear operation is requested during an active perform-
| ance explorer collection session.
| CLRPIN *BASIC Clear Pin. An event is generated whenever a storage
| clear and pin operation is requested during an active
| performance explorer collection session.
| EXCHCLR *BASIC Exchange Clear. An event is generated whenever a
| storage exchange and clear operation is requested
| during an active performance explorer collection
| session.
| READ *BASIC, *DSKIO2 Read. An event is generated whenever a storage
| read operartion is requested during an active perform-
| ance explorer collection session.
| READASYNC *BASIC, *DSKIO2 Read Asynchronous. An event is generated whenever
| a storage asynchronous read operartion is requested
| during an active performance explorer collection
| session.
| READASYNCCALL *BASIC, *DSKIO2 Read Asynchronous Track Call. An event is gener-
| ated whenever the storage manangement read asyn-
| chronous tracked by a call function is accessed
| during an active performance explorer collection
| session.
| READPIN *BASIC, *DSKIO2 Read Pin. An event is generated whenever a storage
| read and pin operation is requested during an active
| performance explorer collection session.
| EXCHREAD *BASIC, *DSKIO2, Exchange Read. An event is generated whenever a
| *FILEOPEN storage exchange read operation is requested during
| an active performance explorer collection session.
| EXCHREADASYNC *BASIC, *DSKIO2 Exchange Read Asynchronous. An event is generated
| whenever a storage asynchronous exchange read
| operation is requested during an active performance
| explorer collection session.
| EXCHREAD- *BASIC, *DSKIO2 Exchange Read Asynchronous Track Call. An event
| ASYNCCALL is generated whenever the storage manangement
| exchange read asynchronous tracked by a call func-
| tion is accessed during an active performance
| explorer collection session.
| WRT *BASIC, *DSKIO2 Write. An event is generated whenever a storage
| write operation is requested during an active perform-
| ance explorer collection session.
| WRTASYNC *BASIC, *DSKIO2 Write Asynchronous. An event is generated whenever
| a storage asynchronous write operation is requested
| during an active performance explorer collection
| session.
| WRTASYNCCALL *BASIC, *DSKIO2 Write Asynchronous Track Call. An event is gener-
| ated whenever the storage manangement write asyn-
| chronous tracked by a call function is accessed
| during an active performance explorer collection
| session.

Chapter 11. Performance Explorer 11-29

| Table 11-8 (Page 2 of 2). Segment Address Register Events
| Event TRCTYPE Description
| WRTASYNCMSGQ *BASIC, *DSKIO2 Write Asynchronous Track Message Queue. An event
| is generated whenever the storage manangement
| write asynchronous function is accessed during an
| active performance explorer collection session.
| WRTASYNC- *BASIC, *DSKIO2 Write Asynchronous Track Message QueueLast. An
| MSGQLAST event is generated whenever the storage
| manangement write last asynchronous tracked by a
| call is requested during an active performance
| explorer collection session.
| WRTASYNCTASK *DSKIO2 Write Asynchronous Track Task. An event is gener-
| ated whenever the storage manangement write last
| asynchronous tracked by the current task is
| requested during an active performance explorer col-
| lection session.
| WRTASYNCWAIT Write Asynchronous Track Wait. An event is gener-
| ated whenever the storage manangement write asyn-
| chronous tracked by is requested during an active
| performance explorer collection session.
| WRTRMV *BASIC, *DSKIO2 Write Remove. An event is generated whenever a
| storage write remove operation is requested during an
| active performance explorer collection session.
| WRTPGOUT *BASIC, *DSKIO2 Write For Page Out. An event is generated whenever
| a storage write for page out operation is requested
| during an active performance explorer collection
| session.
| RMV *BASIC Remove. An event is generated whenever a storage
| remove operation is requested during an active per-
| formance explorer collection session.
| UNPIN *BASIC Unpin. An event is generated whenever a storage
| unpin operation is requested during an active per-
| formance explorer collection session.
| UNPINRMV *BASIC Unpin Remove. An event is generated whenever a
| storage unpin remove operation is requested during
| an active performance explorer collection session.
| UNPINWRT *BASIC, *DSKIO2 Unpin Write. An event is generated whenever a
| storage unpin write operation is requested during an
| active performance explorer collection session.
| ENDOK End of SAR event: OK. An event is generated when-
| ever any specified SAR event completes successfully
| during an active performance explorer collection
| session.
| ENDERR End of SAR event: Error. An event is generated
| whenever any specified SAR event completes with an
| error during an active performance explorer collection
| session.
| STRASYNCTASK Start Asynchronous Task. An event is generated
| whenever during an active performance explorer col-
| lection session.

| The DASD server events (give descriptions here)

11-30 Performance Tools V4R2

| The data that is collected for the DASD server events are placed in the
| QAYPETIDX and QAYPEDSRV files.

| Table 11-9. DASD Server Events

| Event TRCTYPE Description
| CTLEND *DSKSVR Control End. An event is generated whenever a
| server request to create or destroy session completes
| during an active performance explorer collection
| session.
| RQSACP *DSKSVR Request Accepted. An event is generated whenever
| a server function request is accepted and will be
| handled by a server task during an active perform-
| ance explorer collection session.
| RQSRJT *DSKSVR Request Rejected. An event is generated whenever a
| server function request is rejected because of an
| error during an active performance explorer collection
| session.
| SVRTASKSTR *DSKSVR Server Task Started. An event is generated whenever
| a server task is created and is ready to process
| requests during an active performance explorer col-
| lection session.
| SVRTASKEND *DSKSVR Server Task Ended. An event is generated whenever
| a server task is ended and will be removed from the
| system during an active performance explorer col-
| lection session.
| RQSRCV *DSKSVR Request Received. An event is generated whenever
| a server task receives a request for work (read or
| write) during an active performance explorer col-
| lection session.
| RQSENDOK *DSKSVR Request Ended OK. An event is generated whenever
| a server task completes the request with no errors
| during an active performance explorer collection
| session.
| RQSENDERR *DSKSVR Request Ended Error. An event is generated when-
| ever a server task encounters an error while proc-
| essing a request during an active performance
| explorer collection session.

| Program bracket events trace the call flow of programs and procedures as well as
| MI complex instructions. These events give information on when programs, proce-
| dures and MI complex instruction are entered into (or called) and exited from (or
| returned).

| The data that is collected for the dasd server events are placed in the QAYPETIDX
| and QAYPEMBRKT files.

| Table 11-10 (Page 1 of 2). Program bracket events

| Event TRCTYPE Description
| MIENTRY *CALLRTN, *FILEOPEN, Entry. An event is generated whenever a program
| *PGMACT or procedure that is enabled for performance col-
| lection is entered or called during an active per-
| formance explorer collection session.

Chapter 11. Performance Explorer 11-31

| Table 11-10 (Page 2 of 2). Program bracket events
| Event TRCTYPE Description
| MIEXIT *CALLRTN, *FILEOPEN, Exit. An event is generated whenever a program
| *PGMACT or procedure that is enabled for performance col-
| lection exits or returns during an active perform-
| ance explorer collection session.
| MISTR *CALLRTN, *PGMACT Machine Interface Instruction Start. An event is
| generated whenever an MIcomplex instruction is
| routed to or from a program during an active per-
| formance explorer collection session.
| MIEND *CALLRTN, *PGMACT Machine Interface Instruction End. An event is
| generated whenever an MI complex instruction
| completes and returns to the program during an
| active performance explorer collection session.
| MIPRECALL Pre Call. An event is generated whenever a
| program or procedure that is enabled for perform-
| ance collection calls externally to another program
| or procedure during an active performance
| explorer collection session.
| MIPOSTCALL Post Call. An event is generated whenever a
| program or procedure that is enabled for perform-
| ance collection has been returned to from an
| external call to another program or procedure
| during an active performance explorer collection
| session.
| JVAENTRY Java Entry. An event is generated whenever a
| Java method is called during an active perform-
| ance explorer collection session.
| JVAEXIT Java Exit. An event is generated whenever a Java
| method is returned from during an active perform-
| ance explorer collection session.
| JVAPRECALL Java Precall. An event is generated whenever a
| Java method is about to call out some other func-
| tion during an active performance explorer col-
| lection session.
| JVAPOSTCALL Java Postcall. An event is generated whenever a
| Java method is returned from some other function
| during an active performance explorer collection
| session.
| JVANTVMTHSTR Java Native Method Start. An event is generated
| whenever a Java method calls an AS/400 method
| during an active performance explorer collection
| session.
| JVANTVMTHEND Java Native Method End. An event is generated
| whenever a Java method returns from an AS/400
| method during an active performance explorer
| collection session.

| Java events describe activity occurring in the Java virtual machine of the AS/400
| system.

| The data that is collected for the DASD server events are placed in the
| QAYPETIDX and QAYPEJAVA files.

11-32 Performance Tools V4R2

| Table 11-11 (Page 1 of 2). Java events
| Event TRCTYPE Description
| OBJCRT Java object create. An event is generated whenever a
| Java object instance is created during an active per-
| formance explorer collection session.
| GBGCOL Java garbage collection. An event is generated when-
| ever Java is doing garbage collection during an active
| performance explorer collection session.
| GBGCOLSWEEP Java garbage collection sweep. An event is generated
| whenever the garbage collection sweep completes
| during an active performance explorer collection
| session.
| THDCRT Java thread create. An event is generated whenever a
| Java thread is created during an active performance
| explorer collection session.
| THDDLT Java thread delete. An event is generated whenever a
| Java thread is destructed during an active perform-
| ance explorer collection session.
| THDSSP Java thread suspend. An event is generated when-
| ever a Java thread or thread group suspend is started
| during an active performance explorer collection
| session.
| THDRSM Java thread resume - An event is generated when-
| ever a java thread or thread group is resumed during
| an active performance explorer collection session.
| THDWAIT Java thread wai. An event is generated whenever an
| Object::wait() used for thread synchronization is
| started during an active performance explorer col-
| lection session.
| THDNFY Java thread notify. An event is generated whenever
| an Object::notify() used for thread synchronization is
| started during an active performance explorer col-
| lection session.
| THDNFYALL Java thread notify all. An event is generated when-
| ever an Object::notifyAll() used for thread synchroniza-
| tion is started during an active performance explorer
| collection session.
| THDSTTCHG Java thread state change. An event is generated
| whenever a method that causes a thread state transi-
| tion (for example, a thread block) is started during an
| active performance explorer collection session.
| CLSLOAD Java class load. An event is generated whenever a
| Java class is loaded during an active performance
| explorer collection session.
| CLSUNLOAD Java class unload. An event is generated whenever a
| Java class is unloaded during an active performance
| explorer collection session.
| LIBOPR Java class library operation. An event is generated
| whenever a Java native method library is loaded or
| unloaded during an active performance explorer col-
| lection session.

Chapter 11. Performance Explorer 11-33

| Table 11-11 (Page 2 of 2). Java events
| Event TRCTYPE Description
| TFMSTR Java transform start. An event is generated whenever
| a Java program object generation step is started
| during an active performance explorer collection
| session.
| LCKSTR Java lock. An event is generated whenever a synchro-
| nize lock is requested during an active performance
| explorer collection session.
| UNLCK Java unlock. An event is generated whenever a syn-
| chronize lock is released during an active perform-
| ance explorer collection session.
| JVAEND Java end. An event is generated whenever a previous
| Java event is completed during an active performance
| explorer collection session.

| Table 11-12. Selected MI complex instructions

| MI INST TRCTYPE Description
| *REQIO *BASIC Request IO complex MI
| *SNSEXCPD *BASIC Sense exception description
| *DEQ *BASIC Dequeue
| *ENQ *BASIC Enqueue
| *WAITEVT *BASIC Wait on event
| *RETEVTD *BASIC Retrieve event data
| *SUSPR *BASIC Suspend process
| *TERMPR *BASIC Terminate process
| *INITPR *BASIC Initiate process
| *RESPR *BASIC Resume process
| *SIGEVT *BASIC Signal event
| *UNLOCKSL *BASIC Unlock space location
| *LOCK *BASIC Lock object
| *LOCKSL *BASIC Lock space location
| *UNLOCK *BASIC Unlock object
| *DEQWAIT *BASIC Dequeue wait
| *DEACTPG *BASIC, *PGMACT De-activate program
| *DESAGP *BASIC, *PGMACT Destroy activation group

| Basic Report
| The Basic Report provides summary information that includes the definition, run,
| and task information sections for any of the previous types.

11-34 Performance Tools V4R2

 Mapping OPM High-Level Language (HLL) Statements to Source Code
Non-ILE compiled original program model (OPM) program HLL statements that
appear in the PRTPEXRPT *PROFILE report do not map the code source state-
ments. To do the mapping, these programs need to be compiled with the *LIST
generation option.
Note: This mapping only applies to the OPM compilers.

The HLL statements that appear in the report have to be converted to hexadecimal
and matched up with the INST column in the Generated Output section of the
compile listing. The values under the Break column of the same section map the
source statement numbers of the program.

For example, the following CL program results in a partial listing as shown in

Figure 11-12 on page 11-36:
CRTCLPGM PGM(QGPL/CLEXAMPLE)
SRCFILE(QGPL/QCLSRC) GENOPT(\LIST)

| To determine the actual source code statement numbers, do the following steps:
| 1. Compile the original program model (OPM) program with an *LIST generation
| option. This listing includes the original HLL source statement numbers and the
| corresponding MI instructions that were generated for this HLL statement.
| These MI instructions are assigned their own INSTruction number on the listing.
| 2. Collect data that includes the OPM program.
| 3. Print the report by specifying:
| PRTPEXRPT TYPE(\PROFILE) PROFILEOPT(\SAMPLECOUNT \STATEMENT)
| 4. Find the highest count statement number for the OPM program and convert the
| number to its corresponding hexadecimal value. For example, decimal 241 is
| hexadecimal F1 ((F15 *16) + 1.
| 5. Scan the MI statement portion of the listing (Generate Output section) to find
| the hexadecimal instruction number under the column heading INST. On the
| right side of that same print line you see the HLL source statement number
| under the Break column heading. Two lines before the matched INST line you
| see BRK 'HLL source statement number'.
| 6. Find that statement number in your original source portion of the listing.

Chapter 11. Performance Explorer 11-35

 Control Language Source
SEQNBR \...+... 1 ...+... 2 ...+... 3 ...+... 4 ...+... 5 ...+... 6 ...+... 7 ...+... 8 ...+... 9 ...+. DATE
1ðð- /\------------------------------------------------------------------\/
2ðð- /\ Program : LOOPCL \/ ð8/27/95
3ðð- /\ \/
4ðð- /\ Example invocation : \/
5ðð- /\ \/
6ðð- /\ CALL QGPL/LOOPCL \/ ð8/27/95
7ðð- /\ \/
8ðð- /\ \/
9ðð- /\ \/
1ððð- /\ Parameters : none \/
11ðð- /\------------------------------------------------------------------\/
12ðð-
13ðð- PGM:
14ðð- DCL &LOOPCNT \DEC LEN(5 ð) VALUE(1ðððð) ð8/12/95
15ðð- DCL &VAR1 \DEC LEN(5 ð) VALUE(ð) ð8/27/95
16ðð- /\------------------------------------------------------------------\/
17ðð- /\ SIMPLE LOOP WITH SOME MATH COMPUTATIONS \/ ð8/27/95
18ðð- /\ \/ ð8/27/95
19ðð- /\------------------------------------------------------------------\/
2ððð- LOOP: ð7/28/95
21ðð- IF COND(&LOOPCNT \NE ð) THEN(DO) ð8/23/95
22ðð- CHGVAR VAR(&VAR1); VALUE(&LOOPCNT \ &VAR1); ð8/27/95
23ðð- CHGVAR VAR(&VAR1); VALUE(&VAR1 / &LOOPCNT); ð8/27/95
24ðð- CHGVAR &LOOPCNT (&LOOPCNT - 1) ð7/28/95
25ðð- GOTO LOOP ð7/28/95
26ðð- ENDDO ð7/28/95
27ðð- END:
\ \ \ \ \ E N D O F S O U R C E \ \ \ \ \

Figure 11-12 (Part 1 of 2). Mapping Control Language Source to Statement Numbers

11-36 Performance Tools V4R2

5716SS1 V3R6Mð 95ð929 Generated Output ð8/27/95 11:25:31 Page 3
SEQ INST Offset Generated Code \... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ... ... 8 Break
ðððð1 ENTRY \ EXT ;
ðððð2 ððð1 ððððð4 2132 ðð25 ðð26 CPYBWP ?WCLSEPT@,?WWCBSEPT ;
ðððð3 ððð2 ðððððA ð132 ðð21 4ð27 213A CPYBWP ?QCLCLNUP,?WWLISEPT(ðð314) ;
ðððð4 ððð3 ðððð12 ð252 ðð21 ðð4A SETIEXIT ?QCLCLNUP,?WCLRARGLST ;
ðððð5 ððð4 ðððð18 ð283 4ð27 213C ðð4A CALLX ?WWLISEPT(ðð316),?WCLRARGLST,\ ;
ðððð
ðððð6 DCL DD ?CLPVARS(ððððððð7) CHAR(1) AUTO ;
ðððð7 BRK 'PGM ' ; PGM
ðððð8 PGM : ; PGM
ðððð9 BRK 'LOOP ' ; LOOP
ððð1ð LOOP : ; LOOP
ððð11 ððð5 ðððð22 23EF ðð1A ðð18 2ðð1 MODEXCPD ?FCEXCMON,?EMEMONAT,X'ð1' ; LOOP
ððð12 BRK '21ðð ' ; 21ðð
ððð13 ?RCLBLðððð1: ; 21ðð
ððð14 ððð6 ðððð2A 3ð42 ðð31 2ðð1 CPYNV ?WCLCSREI ,ðððð1 ; 21ðð
ððð15 ððð7 ðððð3ð 1846 Cððð ðð13 ððð6 CMPNV(I) &LOOPCNT,P'+ð'/NEQ(?4TEMPððð1); ; 21ðð
ðð16 21ðð
ððð16 ððð8 ðððð3A 1CC2 Cððð ðð16 2ðF1 CMPBLA(B) ?4TEMPððð1 ,C'1'/NEQ(?FLðððð1) ; 21ðð
ðð1ð 21ðð
ððð17 BRK '22ðð ' ; 22ðð
ððð18 ?RCLBLðððð2: ; 22ðð
ððð19 ððð9 ðððð44 3ð42 ðð31 2ðð2 CPYNV ?WCLCSREI ,ðððð2 ; 22ðð
ððð2ð ðððA ðððð4A 1ð4B ðð17 ðð13 ðð14 MULT ?4TEMPððð2,&LOOPCNT,&VAR1 ; 22ðð
ððð21 ðððB ðððð52 1ð42 ðð14 ðð17 CPYNV &VAR1 ,?4TEMPððð2 ; 22ðð
ððð22 ?ICLBLðððð2: ; 22ðð
ððð23 BRK '23ðð ' ; 23ðð
ððð24 ?RCLBLðððð3: ; 23ðð
ððð25 ðððC ðððð58 3ð42 ðð31 2ðð3 CPYNV ?WCLCSREI ,ðððð3 ; 23ðð
ððð26 ðððD ðððð5E 1ð4F ðð17 ðð14 ðð13 DIV ?4TEMPððð2,&VAR1,&LOOPCNT ; 23ðð
ððð27 ðððE ðððð66 1ð42 ðð14 ðð17 CPYNV &VAR1 ,?4TEMPððð2 ; 23ðð
ððð28 ?ICLBLðððð3: ; 23ðð
ððð29 BRK '24ðð ' ; 24ðð
ððð3ð ?RCLBLðððð4: ; 24ðð
ððð31 ðððF ðððð6C 3ð42 ðð31 2ðð4 CPYNV ?WCLCSREI ,ðððð4 ; 24ðð
ððð32 ðð1ð ðððð72 1ð47 ðð17 ðð13 ðððC SUBN ?4TEMPððð2,&LOOPCNT,P'+1'; ; 24ðð
ððð33 ðð11 ðððð7A 1ð42 ðð13 ðð17 CPYNV &LOOPCNT ,?4TEMPððð2 ; 24ðð
ððð34 ?ICLBLðððð4: ; 24ðð
ððð35 BRK '25ðð ' ; 25ðð
ððð36 ðð12 ðððð8ð 3ð11 ððð4 B LOOP ; 25ðð
ððð37 BRK '26ðð ' ; 26ðð
ððð38 BRK 'END ' ; END
ððð39 END : ; END
ððð4ð BRK '27ðð ' ; 27ðð
ððð41 ?ICLBLðððð1: ; 27ðð

Figure 11-12 (Part 2 of 2). Mapping Control Language Source to Statement Numbers

Chapter 11. Performance Explorer 11-37

11-38 Performance Tools V4R2
Chapter 12. Managing the Performance Tools Configuration
For the Manager feature, if you choose the Configure and manage tools option on
the IBM Performance Tools menu, the Configure and Manage Tools display
appears.

à ð
Configure and Manage Tools

Select one of the following:

1. Work with functional areas

2. Delete performance data
3. Copy performance data
4. Convert performance data

Selection or command
===> _____________________________________________________________________
__________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ
From this display you can manage or change the objects used in the Performance
Tools.

For the Agent feature, choose option 2 (Manage performance data) on the IBM
Performance Tools menu.

 Copyright IBM Corp. 1998 12-1

 à ð
IBM Performance Tools for AS/4ðð

Select one of the following:

1. Collect performance data

2. Manage Performance Data
3. Work with Historical Data
4. Advisor

Selection or command
===> _____________________________________________________________________
__________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant
F16=System main menu

á ñ
The Manage Performance Data display will appear.

à ð
Manage Performance Data

Select one of the following:

1. Delete performance data

2. Copy performance data
3. Convert performance data

Selection or command
===> _____________________________________________________________________
__________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel
á ñ
From this display you can manage the objects used in the performance tools.

Work with Functional Areas—Manager Feature

Functional areas provide a way to define and save selection values that you use on
the System and Component Reports. For example, you might save a set of jobs or
users as a functional area. Then each time you use the Print System Report
(PRTSYSRPT) and Print Component Report (PRTCPTRPT) commands, you
specify the name of the functional area to use. Functional areas also work with the
Print Job Report (PRTJOBRPT), Print Pool Report (PRTPOLRPT), Print Trans-
action Report (PRTTNSRPT), and Display Performance Graph (DSPPFRGPH)

12-2 Performance Tools V4R2

 commands. Specify these names on the select functional areas (SLTFCNARA) and
the omit functional areas (OMTFCNARA) parameters.

If you choose option 1 (Work with functional areas) on the Configure and Manage
Tools display, the Work with Functional Areas display appears.

à ð
Work with Functional Areas

Library . . . . . . . . . . . QPFRDATA

Type options, press Enter.

1=Create 2=Change 3=Copy 4=Delete

Functional
Option Area Text
1 My Func Area My department
ACCOUNTING Func Area for Accounting
DEPARTMENT A Func Area for Dept. A
DEPARTMENT B Func Area for Dept. B
DEPARTMENT C Func Area for Dept. C
DEPARTMENT D Func Area for Dept. D
MANAGEMENT Func Area for Managers
MANUFACTURING Func Area for Manufacturing
PAYROLL Func Area for Payroll
SALES FORCE Func Area for Sales Force
SECRETARIAL Func Area for Secretaries
More...
F3=Exit F5=Refresh F12=Cancel F16=Sort by text

á ñ
This display shows the functional areas that exist in the library you specified. To
create a new functional area, type option 1, the name, and the description on the
first line under the Functional Area and Text columns, and press the Enter key. To
select an existing functional area, type a 2 (Change), 3 (Copy), or a 4 (Delete) in
the Option column next to the functional area of your choice.

Creating a Functional Area—Manager Feature

If you choose to create a functional area, the Create Functional Area display
appears.

Chapter 12. Managing the Performance Tools Configuration 12-3

 à ð
Create Functional Area

Functional Area . . . . . . : MY FUNC AREA

Type options, press Enter.

1=Select

Option Job User Option Job User

_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________
_ __________ __________ _ _________ __________

More...
F3=Exit F12=Cancel

á ñ
On this display you specify the job name and the user ID (or both) you want to
include in the functional area. If you choose, you can specify only a job name, only
a user ID, a generic job name (of the form yyy*), or a generic user name. Thus,
WS* in the Job column would include all jobs that have a job name starting with
WS and any user ID name. You can leave a blank field in the Job or User column
to include all jobs with any job name or all jobs with any user ID name.

When you have finished entering all the job names and user IDs, press the Enter
key to create the functional area. Make sure you put a 1 in front of each job and
user you enter.

Changing a Functional Area—Manager Feature

If you choose to change a functional area, the Change Functional Area display
appears.

12-4 Performance Tools V4R2

 à ð
Change Functional Area

Functional Area . : DEPARTMENT A

Text . . . . . . . Func Area for Dept. A

Type options, press Enter.

1=Select 4=Delete

Option Job User Option Job User

1 JOB1 MARY
1 OPGMR
1 DSPð2 A\
1 M\
_ _________ __________ _ _________ _________
_ _________ __________ _ _________ _________
_ _________ __________ _ _________ _________
More...
F3=Exit F5=Refresh F12=Cancel F15=Sort by job name
F16=Sort by user name

á ñ
On this display you specify new job names and user IDs to include in the functional
area by using option 1, or remove jobs and users from the functional area by using
option 4. When you have made all of your entries, press the Enter key to change
the functional area.

Deleting a Functional Area—Manager Feature

If you choose to delete a functional area, the Confirm Delete of Functional Areas
display appears, listing the functional areas you selected for deletion. Press the
Enter key to delete them.

Copying a Functional Area

If you choose to copy a functional area, the Copy Functional Area (CPYFCNARA)
command prompt appears. Fill in the prompts and press the Enter key to copy the
functional area.

Delete Performance Data

Use the Delete performance data option on the Configure and Manage Tools
display to delete performance data that you no longer need on your system. When
you choose option 2, the Delete Performance Data display appears.

Chapter 12. Managing the Performance Tools Configuration 12-5

 à ð
Delete Performance Data

Library . . . . . . QPFRDATA__

Type option, press Enter.

4=Delete
Option Member Text Date Time
_ XYZ 12/15/95 14:ð5:55
_ PERFTESTC4 2 hours w/ 5 minute intervals 12/15/95 ð8:ð5:48
_ PERFTESTC3 Duration of 2 hours 12/14/95 ð9:21:44
_ PERFTESTC2 12/11/95 14:42:46

Bottom
F3=Exit F5=Refresh F12=Cancel F15=Sort by member F16=Sort by text

á ñ
The members that appear on this display are those used on the Start Performance
Monitor (STRPFRMON) command for the keyword MBR when data was collected.
To delete a member from this list, type a 4 (Delete) next to the appropriate member
and press the Enter key. The member you delete is deleted from the following data
collection files:

QAPGSUMD QAPMIOPD
QAPMAPPN QAPMPOOL
QAPMASYN QAPMRESP
QAPMBSC QAPMRWS
QAPMBUS QAPMSAP
QAPMCIOP QAPMSBSD
QAPMCONF QAPMSNA
QAPMDBMON QAPMSNADS
QAPMDDI QAPMSTNE
QAPMDIOP QAPMSTNL
QAPMDISK QAPMSYS
QAPMDMPT QAPMTJOB
QAPMECL QAPMTSK
QAPMETH QAPMX25
QAPMFRLY QAPTLCKD
QAPMHDLC QTRIDX
QAPMHDWR QTRDMPT
QAPMIDLC QTRJOBT
QAPMJOBS QTRJSUM
QAPMLAPD QTRSLWT
QAPMLIOP QAPMSTND
QAPMMIOP QAPMSTNY
QTRTSUM

12-6 Performance Tools V4R2

Copy Performance Data
Use the Copy performance data option on the Configure and Manage Tools display
to make copies of performance data members. When you choose option 3, the
Select Performance Member display appears.

à ð
Select Performance Member

Library . . . . . . QPFRDATA__

Type option, press Enter.

1=Select

Option Member Text Date Time

_ FRIDAY Performance Data for Friday 1ð/27/95 1ð:ð5:46
_ THURSDAY Performance Data for Thursday 1ð/26/95 12:ðð:34
1 WEDNESDAY Performance Data for Wednesday 1ð/25/95 13:5ð:15
1 TUESDAY Performance Data for Tuesday 1ð/24/95 13:55:ð8
1 MONDAY Performance Data for Monday 1ð/23/95 16:25:39
_ TESTRUN Test run of system 1ð/19/95 2ð:31:42
_ Q95291ð958 1ð/18/95 ð9:58:45
_ Q9529ð2ðð9 1ð/17/95 2ð:ð9:23

F3=Exit F12=Cancel F15=Sort by member F16=Sort by text

F19=Sort by date/time

á ñ
The members that appear on this display are those used on the Start Performance
Monitor (STRPFRMON) command for the keyword MBR when data was collected.
To copy a member or members from the list, type a 1 (Select) next to the appro-
priate member(s) and press the Enter key. The Copy Performance Data display
appears.

Chapter 12. Managing the Performance Tools Configuration 12-7

 à ð
Copy Performance Data Member

Type choices, press Enter.

--------Copy From-------- --------Copy To--------

Member Library Member Library
MONDAY QPFRDATA MONDAY NEWLIB
TUESDAY QPFRDATA TUESDAY NEWLIB
WEDNESDAY QPFRDATA WEDNESDAY NEWLIB
Bottom

F3=Exit F12=Cancel

á ñ
This display shows you the members you selected to copy and where they are to
be copied to. For each member listed, type the name of the new member and the
library that contains it in the Copy To entries of the display, and then press the
Enter key. When the copy completes, you have exact copies of the old perform-
ance members in the new performance members for the following files:

QAPMAPPN QAPMLAPD
QAPMASYN QAPMLIOP
QAPMBSC QAPMMIOP
QAPMBUS QAPMPOOL
QAPMCIOP QAPMRESP
QAPMCONF QAPMRWS
QAPMDBMON QAPMSAP
QAPMDDI QAPMSBSD
QAPMDIOP QAPMSNA
QAPMDISK QAPMSNADS
QAPMDMPT QAPMSTND
QAPMECL QAPMSTNE
QAPMETH QAPMSTNL
QAPMFRLY QAPMSTNY
QAPMHDLC QAPMSYS
QAPMHDWR QAPMTJOB
QAPMIDLC QAPMTSK
QAPMIOPD QAPMX25
QAPMJOBS

Convert Performance Data (CVTPFRDTA) Command

Use the Convert performance data option on the Configure and Manage Tools
display to convert performance data. The data is converted to the file formats
needed to be processed by the current release of the performance
measurement/analysis tools.

12-8 Performance Tools V4R2

à ð
Configure and Manage Tools

Select one of the following:

1. Work with functional areas

2. Delete performance data
3. Copy performance data
4. Convert performance data

Selection or command
===> 4________________________________________________________________________
______________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ
When you select option 4, the Convert Performance Data (CVTPFRDTA) command
prompt display appears.

à ð
Convert Performance Data (CVTPFRDTA)

Type choices, press Enter.

From library . . . . . . . . . . ___________ Name

To library . . . . . . . . . . . ___________ Name
Job Description . . . . . . . . \USRPRF____ Name, \USRPRF, \NONE
Library . . . . . . . . . . . ___________ Name, \LIBL, \CURLIB

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

á ñ
The Convert Performance Data (CVTPFRDTA) command converts performance
data from the previous release to the formats needed to be processed by the
current release of the performance measurement/analysis tools. First, the release
level on which the data was collected is determined. Then, all members of all files
that need conversion are converted to the appropriate format.

The following files must be present for the conversion to take place:

Chapter 12. Managing the Performance Tools Configuration 12-9

 QAPMCIOP QAPMLIOP
QAPMCONF QAPMPOOL
QAPMDIOP QAPMRESP
QAPMDISK QAPMSYS
QAPMJOBS

The following files are copied, or converted if necessary, if they are present:

QACPCNFG QAPMDMPT
QACPGPHF QAPMECL
QACPPROF QAPMETH
QACPRESP QAPMHDLC
QAITMON QAPMIDLC
QAPGHSTD QAPMLAPD
QAPGHSTI QAPMMIOP
QAPGPKGF QAPMSBSD
QAPMASYN QAPMTSK
QAPMBSC QAPMX25
QAPMBUS QAPTAPGP

The conversion can be done in the library in which the current data resides or in a
different library. If the conversion is done in the same library, the current data is
replaced by the new data. If the conversion is done in a different library, the new
data exists in the new library while the current data continues to exist in the current
library.

Notes:
1. If a different library is specified for the new data, those files in the current
library that do not need conversion are copied to the new library.
2. If user-created logical files exist over any of these files, you must delete and
re-create these logical files after the performance data has been converted.
3. Historical data cannot be converted without the performance monitor files
required for converting.

To convert performance data collected prior to the current release, complete the
following items on the display.
From library Specifies the library that contains the data being converted.
To library Specifies the library that contains the converted data.
Job Description Specifies the job description used to submit the file-conversion job
for batch processing.

The possible job description values are:

*USRPRF The job description defined for the submitting job’s user profile.
job-description-name Specify the name of the job description to be used.
*NONE A batch job is not submitted. Processing continues interactively while the
user waits.

The possible library values are:

*LIBL The library list is used to locate the job description.

12-10 Performance Tools V4R2

*CURLIB The current library for the job is used to locate the job description. If no
current library entry exists in the library list, QGPL is used.
library-name The library where the job description is located.
Note: If the conversion takes place interactively, the user’s work station is not
available for other use during this time, which can be significant for long
jobs.

Chapter 12. Managing the Performance Tools Configuration 12-11

12-12 Performance Tools V4R2
Chapter 13. A Problem Analysis Case Study
This chapter provides users of the AS/400 system with an initial approach to deter-
mining the source of performance problems using available system tools (both
operating system functions and additional licensed programs).

This example provides an introduction to performance analysis and shows you

some performance analysis techniques. The scenario describes how a user who is
experienced in performance analysis assists a company in resolving a performance
concern. You can see how the expert identifies the real problem, isolates the
cause, and provides the recommended solution.

You should read “Performance Analysis Overview” on page 1-8 before reading this
case study.

Some of the tools discussed are not available in the Agent feature. Appendix D,
Comparison of Performance Tools, provides additional information about Perform-
ance Tools functions.
Note: Although the technique used represents just one of many different
approaches to performance analysis, and the problem described is a small
subset of the real-life possibilities, the example is designed to provide initial
guidance in developing an overall strategy for performance problem anal-
ysis. The names of people and the events described in this chapter are ficti-
tious, and any likeness to actual people is purely coincidental. Because
customer applications and requirements vary, IBM makes no representation
or warranty that the methodology described herein will solve or eliminate
unique customer performance problems.

Introduction to Performance Analysis

Performance problem analysis is a methodology for investigating, measuring, and
correcting deficiencies so that system performance meets the user’s expectations. It
does not matter much that the “system” is a computer; it could be an automobile or
a washing machine. The problem-solving approach is essentially the same:
1. Understand the symptoms of the problem.
2. Use tools to measure and define the problem.
3. Isolate the cause.
4. Correct the problem.
5. Use tools to verify that the problem is corrected.

Initially, the analyst knows the user is not satisfied with the way the system is
working. For example, it may be running too slow, too noisy, too hot, and so on.
The analyst, mechanic, or repair person must first understand what the problem
really is. The best way to find out is to observe the problem condition personally.
Can the analyst confirm the user’s complaint? If the analyst cannot, he should get
as much information as possible from those users who have experienced the
problem. What are the most common problem descriptions?

The key to success with any performance issue is to have a clear definition of the
users’ performance criteria. In other words, given the application mix, what do the

 Copyright IBM Corp. 1998 13-1

 users want from the system in terms of interactive response time, batch throughput,
and processing requirements? For example, a system that supports an interactive
order-entry application may have a response time criterion to ensure that customers
do not perceive abnormal delays. Another criterion may require that end-of-day
processing be completed by a specific time. Given these requirements, the analyst
can establish performance objectives around system resource utilization guidelines.
With a clear statement of goals and objectives, performance analysis can proceed
on a firm basis.

When the objectives are understood, it is important to assess whether the hardware
configuration is adequate to support the workload. Is there enough processing unit
capacity? Is main storage sufficient for the application mix? It the analyst answers
these questions first, perhaps through capacity planning modeling techniques,
needless effort can be avoided later.

With an understanding of the symptoms of the problem and the objectives to be

met, the analyst can formulate a hypothesis that may explain the cause of the
problem. The analyst can use certain OS/400 commands and Performance Tools to
measure the system performance. The analyst should review the measured data to
further define the problem and to validate or reject the hypothesis. When the
apparent cause or causes have been isolated, a solution can be proposed. The
analyst should deal with one solution at a time. Then changes can be made and
tested. Again, the analyst’s tools can, in many cases, measure the effectiveness of
the solution and look for possible side effects.

To achieve optimum performance, one must recognize the interrelationship

between the critical system resources and attempt to balance these resources,
namely processing unit, disk, main storage, and, for communications, remote lines.
Each of these resources may become a performance problem.

Improvements to system performance, whether to interactive throughput, interactive

response time, batch throughput, or some combination, may take many forms from
simply adjusting activity level or pool size to changing the application code itself.
Ultimately, however, any improvement will come only through analysis of the critical
resources (processing unit, main storage, disk, and remote lines) and contention for
system and application objects.

The Case Study

This scenario starts with a company called Armstrong Sporting Goods, Inc. (a ficti-
tious company). As a distributor of sporting equipment throughout the southeastern
United States, Armstrong selected the AS/400 system as a means for automating
much of their order-entry, accounting, inventory, and shipping operations. High
quality customer service is critical to the continued success of this company.

The Players
Sue Miller is the new data processing (DP) manager for Armstrong and is the
person who provided the IBM support team with most of the information regarding
the perceived performance problem. Having been the DP manager for just a short
time, Sue is anxious to establish her credibility with the company by quickly
addressing an end-user concern over system performance.

13-2 Performance Tools V4R2

 Bob Williams is the assigned IBM systems engineer. He has been asked by Sue to
assist the DP staff in resolving the performance issue. In this scenario, he is the
expert.

As you read through the rest of this story, you can look over Bob’s shoulder as he
observes the symptoms of Armstrong’s performance problem and then proceeds to
isolate the cause. Along the way, you will be introduced to additional people who
prove to be instrumental in getting the problem resolved.

The Configuration
Here is the configuration of the system in this story:
Ÿ One AS/400 Model 510, 384 megabytes main storage
Ÿ Two communications lines
– A 2400 baud electronic customer support switched line
– A spare line that is currently unused
Ÿ One 6380 Tape Unit
Ÿ One 4028 Printer
Ÿ Twenty-four 3197 Display Stations
Ÿ Two 4224 Printers
Ÿ Four 6603 Disk Units
Ÿ One 6607 Disk Unit
| Ÿ OS/400 Version 4 Release 2 licensed program

With this information as the base for our example, let’s begin!

The Problem
It was 9:00 a.m. Monday morning, and Sue Miller had just finished introducing
herself as the new data processing manager to Bob Williams, a systems engineer
from IBM. The two of them were in her office to review the systems management
procedures currently in place at Armstrong. During the discussion, Sue mentioned
that no formal attempt was being made to monitor the system’s performance on a
regular basis. Other activities, such as system backup and change management,
had already been addressed by Armstrong, but Sue felt that she needed to have a
better understanding of how well their current system was handling the daily
demands of the company. This interest was actually prompted by concerns brought
up at a recent meeting with the department managers. Some of the end users had
complained that the system was running too slow and at times appeared to “go to
sleep.”

Bob was happy to hear that Sue wanted to start developing a performance man-
agement strategy for Armstrong. He remarked that he had worked with several
companies in the past who unfortunately waited until a serious situation occurred
before starting to make an effort to better understand their system requirements.
Without historical information to compare past performance, the problem analysis
became much more difficult.

Bob and Sue then continued talking the rest of the morning about other topics of
interest involving the data processing department. At the close of their discussion,

Chapter 13. A Problem Analysis Case Study 13-3

 Bob suggested another meeting to further investigate the source of the perform-
ance concerns. In the meantime, Sue was asked to do the following:
1. Read the Work Management book to better understand performance guidelines
and basic tuning techniques. This would help prepare Sue for the upcoming
analysis activities that Bob would assist her with.
2. Use the error reporting functions, such as the Print Error Log (PRTERRLOG)
command, to see if the system is experiencing hardware problems. Although
this should be a part of normal systems management, all performance analysis
activities should first ensure that the system is running error free.
3. Install the Performance Tools Manager feature that Armstrong had purchased
two weeks earlier. This would assist them in their investigation of the problem.
4. Survey
5. the end users to find out who was experiencing unsatisfactory system perform-
ance and of what type (that is, interactive response time, batch throughput, and
so forth). The performance objectives for those end users should then be deter-
mined and put in writing.

With that, Bob left with the agreement that they would get together on Friday
morning.

Review
Sue is new to Armstrong and is not familiar with the system’s performance history.
She must quickly learn the objectives of the end users. How important is it that
certain display station operators receive subsecond response time, and is it realistic
given the requirements of the application? Are there any critical batch jobs that
must be finished by the end of the day? These kinds of questions need to be
answered for Sue to determine if a problem exists.

As Bob mentioned, even though the AS/400 system provides software tools to
monitor performance, both in the operating system and in optional licensed pro-
grams, many companies do not track their system’s resource usage. Using the
AS/400 system’s ability to continuously collect performance data, a business can
review workload trends on a periodic basis.

At this point, Sue suspects a problem exists because of informal remarks by some
of the end users. She has no solid evidence describing the problem and, therefore,
cannot give Bob any concrete information to work with. We are not even sure if
there is a performance problem. This is usually where many analysis experiences
begin. Bob decided that before proceeding with the analysis, the customer should
first review basic tuning guidelines, make sure an important software product he
relies on (the Performance Tools Manager feature) is ready to use, check to see if
the system is running without hardware problems, and gather more information
from the end users.

Checking the System’s Performance

On Friday morning, Bob returned to Armstrong to begin working with Sue on ana-
lyzing the system’s performance. Two days earlier, Sue met with all of the depart-
ment managers where she asked them to survey their staffs on how well they
thought the system was performing. The results were to be returned to her by the
following Monday. Also, a review of the system error reports did not indicate that
the system was having hardware problems.

13-4 Performance Tools V4R2

Bob felt that the first step in analyzing system performance was to review data from
the system interactively using the control language (CL) commands:
Ÿ Work with System Status (WRKSYSSTS)
Ÿ Work with Active Jobs (WRKACTJOB)
Ÿ Work with Disk Status (WRKDSKSTS)

Using these commands, he could quickly see if the system was able to handle the
requests for processing unit, disk, and main storage adequately at that instant in
time. He cautioned that because the results changed dynamically with the work-
load, he could not determine for sure that the system had all the capacity it needed
all the time. Also, it was important that the time selected to run these commands
did not include work not normally running (for example, excessive sign-ons and
sign-offs). Sue assured Bob that now would be a good time to look at the system.

The following illustrations show the results of the commands and how Bob inter-
preted them. (All output was generated using the OS/400 Version 3 Release 7
licensed program.)

First Bob issued the Work with System Status (WRKSYSSTS) command.

à ð
Work with System Status SYS4ðð
ð7/ð7/98 ð9:31:43
% CPU used . . . . . . . : 55.7 Auxiliary storage:
Elapsed time . . . . . . : ðð:ð9:31 System ASP . . . . . . : 8.12 G
Jobs in system . . . . . : 1ð2 % used . . . . . . . . : 57.5494
% addresses used: Total . . . . . . . . : 8.12 G
Permanent . . . . . . : 2.483 Current unprotect used : 326 M
Temporary . . . . . . : .ð26 Maximum unprotect . . : 328 M

Type pool size and activity level changes, press Enter.

System Pool Reserved Max -----DB----- ---Non-DB---

Pool Size (K) Size (K) Active Fault Pages Fault Pages
1 1ð6564 29356 +++ .ð .ð .2 1.6
2 174544 ð 4 .9 2.9 1.8 4.6
3 5256 ð 4 .ð 1.3 .ð .4
4 1ð6852 ð 12 3.2 27.2 3.9 24.1

Bottom
Command
===>
F3=Exit F4=Prompt F5=Refresh F9=Retrieve F1ð=Restart
F11=Display transition data F12=Cancel F24=More keys

á ñ
Ÿ The overall processing unit use was 55.7 percent and did not reflect an exces-
sively busy system.
Ÿ The elapsed time for measurement was greater than 5 minutes but less than
15 minutes–a good choice when looking for valid data that is not skewed by
short surges of activity or long periods that tend to average out problems.
Ÿ The number of jobs in the system at first appeared high to Sue, but Bob
explained that this number reflected all the jobs the system was keeping track
of, even if they had finished but still had output yet to print (for example, job
logs).

Chapter 13. A Problem Analysis Case Study 13-5

 Ÿ The fault rate of the machine storage pool (always system pool 1) did not
exceed 1 fault per second, indicating that pool 1 was large enough. (More infor-
mation on performance tuning is available in the Work Management book.)
Ÿ The fault rate of the rest of the machine storage pools (system pools 2 through
4) was not too heavy ( database + nondatabase < 10) and the total faults of all
the pools was less than 15. In general, main storage did not appear to be over-
committed.

Bob pressed F11 to select the second view.

à ð
Work with System Status SYS4ðð
ð7/ð7/98 14:ð7:43
% CPU used . . . . . . . : 55.7 Auxiliary storage:
Elapsed time . . . . . . : ðð:ð9:31 System ASP . . . . . . : 8.12 G
Jobs in system . . . . . : 1ð2 % used . . . . . . . . : 57.5494
% addresses used: Total . . . . . . . . : 8.12 G
Permanent . . . . . . : 2.483 Current unprotect used : 326 M
Temporary . . . . . . : .ð26 Maximum unprotect . . : 328 M

Type pool size and activity level changes, press Enter.

System Pool Reserved Max Active-> Wait-> Active->

Pool Size (K) Size (K) Active Wait Inel Inel

1 1ð6564 29356 +++ .ð .ð .ð
2 174544 ð 4 2.5 .ð .ð
3 5256 ð 4 8.3 .ð .ð
4 1ð6852 ð 12 2ð.1 2.ð .ð

Bottom
Command
===>
F3=Exit F4=Prompt F5=Refresh F9=Retrieve F1ð=Restart
F11=Display transition data F12=Cancel F24=More keys

á ñ
Ÿ The ratio of Wait->Ineligible to Active->Wait for the interactive pool (system pool
4) was approximately 10% and confirmed to Bob that the activity level was set
properly. He added that many customers set their activity levels so that the
Wait->Ineligible is always zero. The level could be too high, causing major
problems during exceptionally busy periods.
Ÿ Active->Ineligible for system pool 4 was zero. Usually, any value greater than
zero in the interactive storage pool is a good indication that jobs are exceeding
their time-slice values and may be candidates for submission to batch for proc-
essing.
Bob then issued the Work with Active Jobs (WRKACTJOB) command.

13-6 Performance Tools V4R2

à ð
Work with Active Jobs SYS4ðð
ð7/ð7/98 ð9:47:3ð
CPU %: 57.2 Elapsed time: ðð:ð8:46 Active jobs: 35

Type options, press Enter.

2=Change 3=Hold 4=End 5=Work with 6=Release 7=Display message
8=Work with spooled files 13=Disconnect ...

Opt Subsystem/Job User Type CPU % Function Status

QBATCH QSYS SBS .ð DEQW
ARPOST ACTð7 BCH 24.4 PGM-AR32ð RUN
QCMN QSYS SBS .ð DEQW
QCTL QSYS SBS .ð DEQW
DSPð1 QSYSOPR INT .ð MNU-MAIN DSPW
QINTER QSYS SBS 1.ð DEQW
DSPð2 SHIPð1 INT 1.3 MNU-SMðð1 DSPW
DSPð9 SERVð2 INT .ð CMD-WRKSPLF DSPW
DSP1ð SERVð3 INT .7 PGM=CS11ð RUN
More...
Parameters or command
===>
F3=Exit F5=Refresh F1ð=Restart statistics F11=Display elapsed data
F12=Cancel F23=More options F24=More keys

á ñ

Ÿ The active job count was 35. When divided into the number of jobs in the
system (102/35=3.0), the result showed that Armstrong was doing a good job of
cleaning up the job logs and keeping the number of jobs the system tracked to
a minimum.
Ÿ No interactive jobs were using an excessive amount of processing unit use
(more than 2 percent).
Next, Bob issued the Work with Disk Status (WRKDSKSTS) command.

à ð
Work with Disk Status SYS4ðð
ð7/ð7/98 1ð:ð3:59
Elapsed time: ðð:ð9:11

Size % I/O Request Read Write Read Write %

Unit Type (M) Used Rqs Size (K) Rqs Rqs (K) (K) Busy
1 66ð3 1967 67.1 .7 6.3 .2 .4 8.3 5.3 9
2 66ð3 1967 56.6 1.1 4.5 .2 .8 5.ð 4.4 5
3 66ð7 4194 55.3 1.1 6.5 .5 .6 6.9 6.2 13

Bottom
Command
===>
F3=Exit F5=Refresh F12=Cancel F24=More keys

á ñ

Ÿ Except for disk unit 1 (load source unit), all the other units had approximately
the same amount of space used, indicating an evenly distributed system, and
none of those units were over 75 percent full.

Chapter 13. A Problem Analysis Case Study 13-7

 Ÿ No one drive was busy more than 13 percent of the time, and they were well
under the threshold of 40 percent.

The net result of Bob’s initial observation of the system showed that the system
was responding well to the workload at that moment in time. Sue again commented
that this period of the day was a good representation of Armstrong’s normal
demands on the system.

Bob felt he had a good idea of what the overall system was doing, but he planned
to later validate his findings by using the Advisor option on the Performance Tools
menu. The advisor is a tool that can be run over data gathered by the performance
monitor to provide conclusions and recommendations about system performance. In
the meantime, with the feedback from the different departments still pending, Bob
suggested using another means of gathering performance data from the system.
This could be done by starting the performance monitor using the Work with Per-
formance Collection (WRKPFRCOL) command. With this command, information
similar to that provided by the earlier commands and additional detailed data on job
processing could be collected over a number of days (that is, with multiple data
collections) without operator intervention. Another way to start the performance
monitor would be to use the Start Performance Monitor (STRPFRMON) command,
but this command does not allow for automatic data collection, which is what Bob
and Sue were after. The collected data could then be reviewed through commands
provided by Performance Tools.

To begin the process of collecting performance data with the WRKPFRCOL

command, Bob explained that first they would have to decide when to run the per-
formance monitor and for how long. Without knowing exactly when the problem was
occurring, Bob suggested they collect data for the entire first shift with 15-minute
collection intervals. Using this size interval time would help to identify peak work-
loads that deserved further attention. Any longer intervals might mask a problem.
Because Bob and Sue were still not sure of the type of problem that existed, the
data collection would not include trace information. Tracing is a process by which
information about each job state transition is recorded in a special table. The data
may later be dumped to a database file, which can then be processed by Perform-
ance Tools. Tracing can result in a large amount of collected data, which could
affect system performance when dumped. Normally, a more selective collection
process can be used during problem isolation to limit the amount of data.

Sue entered the WRKPFRCOL command and specified the values as shown in the
following display:

13-8 Performance Tools V4R2

à ð
Work with Performance Collection

Type options, press Enter.

1=Add 2=Change 3=Hold 4=Remove 5=Display 6=Release

Performance
Opt Collection Status Description
1 PERFPROB__

Bottom
F3=Exit F5=Refresh F12=Cancel

á ñ
The following display was shown next. Sue typed the values that are highlighted.

à ð
Add Performance Collection (ADDPFRCOL)

Type choices, press Enter.

Performance Collection . . . . . > PERFPROB Name
Collection days . . . . . . . . \MON \ALL, \MON, \TUE, \WED...
+ for more values \TUE
Collection start time . . . . . ð8ðð HHMM
Text 'description' . . . . . . . All day collection
Library . . . . . . . . . . . . QPFRDATA Name
Text 'description' . . . . . . . Unknown Performance Problem
Time interval (in minutes) . . . 15 5, 1ð, 15, 2ð, 25, 3ð, 35...
Stops data collection . . . . . \ELAPSED \ELAPSED, \TIME
Days from current day . . . . . ð ð-7
Hour . . . . . . . . . . . . . . 1ð ð-168
Minutes . . . . . . . . . . . . ð ð-59
Data type . . . . . . . . . . . \ALL \ALL, \SYS
Trace type . . . . . . . . . . . \NONE \NONE, \ALL
Dump the trace . . . . . . . . . \YES \YES, \NO
Job trace interval . . . . . . . .5 .5 - 9.9 seconds
Job types . . . . . . . . . . . \DFT \NONE, \DFT, \ASJ, \BCH...
+ for more values
More...
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
á ñ
Sue added a collection for Monday and Tuesday from 8:00 a.m until 6:00 p.m. A
collection period of ten hours would include data on sign-offs and batch jobs at the
end of the day. Bob commented to Sue that because this was the first time that
they were adding a performance collection, they would have to submit the perform-
ance collection job after using the ADDPFRCOL command by typing the following
command:

SBMJOB JOBD(QGPL/QPFRCOL) RQSDTA(\JOBD)

This command ensures that the performance monitor is started at the right times.

Chapter 13. A Problem Analysis Case Study 13-9

 Bob left and planned to return at the end of the week to review the output with her.
By that time, Sue should have collected data over two days and had time to go
over the results of her end-user survey.

Review
Bob lacked information on who was experiencing the performance problem, so he
decided to take some preliminary steps in understanding how well the system was
responding to the daily workload. He did this by using the standard system com-
mands, which dynamically show usage of main storage, processing unit, and disk.
The important point that Bob wanted Sue to understand was that these commands
only displayed this information for a very specific point in time and could not be
used to represent the system’s performance under all the different workloads it had
to handle. This was a quick means of looking for obvious resource problems.

The next step was to capture data over a longer period of time using the Work with
Performance Collection (WRKPRFCOL) command. Maybe the problem was occur-
ring at a specific time of the day.

In most everyday situations, performance data could be collected over large periods
of time to get a good idea of system activity and trends. Sampling intervals of
longer duration (20 to 30 minutes) are fine for normal system tracking, but Bob and
Sue are investigating a possible problem. Shorter intervals (10 to 15 minutes)
would help to highlight a problem.

Still critical to Bob’s investigation was the result of Sue’s survey. They still did not
know what kind of a problem they were facing. It is important to thoroughly define
the problem.

Reviewing the End-User Survey Results

On Tuesday, Sue received the final survey results from all the end users. Following
is a copy of the survey form Sue distributed.

13-10 Performance Tools V4R2

 Armstrong Computer End-User Survey

On a scale of 1 to 5, please rate how well the computer system meets your
needs in the following categories:

1 = Excellent, 2 = Satisfactory, 3 = Average,

4 = Needs some improvement, 5 = Needs much improvement

1. Availability of the computer ___

2. Interactive response time ___
3. Timeliness of printed output requests ___
4. Timeliness of batch run requests ___

For those items answered with a 4 or 5, please indicate any concerns

you might have.___________________________________________
________________________________________________________
________________________________________________________

(We will follow up this survey with personal interviews for those
who would like to help the data processing department improve its
services to all the end users.)

Thirty-seven surveys were returned. Sue decided to concentrate on only the

returned forms that indicated a 4 of 5 in any of the categories. She noticed that
only two of the surveys had reflected a dissatisfaction with the system, and both
were from the order-entry department. Also, the only category with negative
responses was number two, Interactive response time. One of the two negative
surveys included the following comment: Ever since the new procedure, which
allowed customers to call in their orders, was put into use, the system seemed to
take a very long time before the entry display appeared.

Sue met briefly with the order-entry department to discuss their survey responses
and to better understand their performance requirements. During the meeting, Sue
learned that the department’s daily workload included both batch and interactive
processing. Their batch jobs ran mostly in the evenings unattended and were not
presenting a problem. The interactive jobs, however, were experiencing much
longer response times than the department’s objective of 2 seconds. Sue reviewed
some basic application requirements, such as the average number of database
read operations per transaction, and could not readily determine the source of the
problem. Sue then decided that it would be better to review her findings with Bob
on Friday.

Analyzing System Performance

On Friday morning, Bob arrived to analyze the collected performance data. First,
Sue updated him on the results of the survey. Bob was very interested in the con-
cerns of the order-entry department and commented that they would investigate the
order-entry application. First, he would like to analyze the system performance once
more using the advisor. Below is the sequence of events and displays that Bob
used to perform the system analysis using the advisor.

Chapter 13. A Problem Analysis Case Study 13-11

 Bob started at the Performance Tools menu and selected the Advisor option. He
then selected the appropriate library and member and pressed the Enter key, which
took him to the Select Time Intervals to Analyze display.

à ð
Select Time Intervals to Analyze

Member . . . . . . . . PERFPROB Library . . . . . . : QPFRDATA

Type options, press Enter

1=Select

---Tns---- --CPU Util--- High Util --Pool Fault/Sec--

Opt Date Time Cnt Rsp Tot Int Bch Dsk Unit Mch Usr ID Excp
_ ð1/15 ð8:15 3ð9 .84 16 1ð 3 1 ððð1 ð ð ð3 192ð
_ ð1/15 ð8:29 266 .46 6 3 1 1 ððð1 ð ð ð3 1ð15
_ ð1/15 ð8:44 635 .87 24 15 5 1 ððð1 ð ð ð3 1174
_ ð1/15 ð8:59 494 .92 53 3ð 15 1 ððð1 ð ð ð3 1229
_ ð1/15 ð9:14 318 .7ð 62 32 2ð 1 ððð1 ð ð ð3 11ð3
_ ð1/15 ð9:29 526 .89 71 4ð 25 1 ððð1 ð ð ð3 1573
_ ð1/15 ð9:44 574 .73 43 2ð 15 1 ððð1 ð ð ð3 1668
_ ð1/15 ð9:59 399 .94 48 2ð 19 1 ððð1 ð ð ð3 135ð
_ ð1/15 1ð:14 243 4.45 11 5 2 1 ððð1 ð ð ð3 192ð
_ ð1/15 1ð:29 246 1.49 24 15 3 1 ððð1 ð ð ð3 1834

Bottom
F3=Exit F5=Refresh F11=Display histogram F12 Cancel F13=Select all

á ñ
Bob suggested to Sue that they analyze all the intervals at this stage to get an idea
of overall system performance. Sue agreed and Bob pressed F13 (Select all) and
pressed the Enter key.

à ð
Display Recommendations

Member . . . . . . . : PERFPROB Library . . . . . . : QPFRDATA

System . . . . . . . : SYS4ðð Version/release . . : 4/2.ð
Start date . . . . . : ð7/ð1/98 Model . . . . . . . : D45
Start time . . . . . : ð8:ðð:ð1 Serial number . . . . : XX-XXXX
Type options, press Enter.
5=Display details
Option Recommendations and conclusions
_ Recommendations
_ Examine error logs for indications of problems.
_ Conclusions
_ Pool 3 fault rate is well below guidelines of 25.ð
_ Pool 4 fault rate is well below guidelines of 25.ð
_ Pool 2 W->I transition zero. Fault rate within guidelines.
_ No performance problems were detected in system data file.
_ No performance problems found on SDLC line MCLINE
_ No performance problems found with DIOP(s)
_ No performance problems found with CIOP(s)
_ Interval Conclusions
_ ASP 1 arm % busy ranged from 21.9% on arm ððð8 to 1ð.2% on arm ððð4.
_ Total system I/O during all selected intervals was 4362ð3 .
More
F3=Exit F6=Print F9=Tune system F12=Cancel F21=Command line
á ñ
The Display Recommendations display showed Bob and Sue that the system was
performing within the guidelines and that no system-related problems or errors
where affecting the performance of the system. After having completed the system

13-12 Performance Tools V4R2

analysis using the advisor, which confirmed his analysis earlier in the week, he
mentioned to Sue that another way to quickly analyze system data and view trends
was to use performance graphics.

Using Performance Graphics—Manager Feature

Following is the sequence of events that Bob specified to produce the graphs. Bob
went to the Performance Tools menu and selected Option 9 (Performance
graphics). Then the following display appeared:

à ð
PERFORMG Performance Tools Graphics
System: SYS4ðð

Select one of the following:

1. Work with graph formats and packages

2. Work with historical data
3. Display graphs and packages

7ð. Related commands

Selection or command
===> 3

F3=Exit F4=Prompt F9=Retrieve F12=Cancel F13=Information Assistant

F16=System main menu

á ñ
Bob explained to Sue that Performance Tools contains numerous pre-formatted
graphs for customers to user. Option 1 allows the user to work with the graph
formats and packages, and option 2 allows the user to create historical data from
data collected over different monitor runs (for example, once a week for a month).
Historical data summarizes performance members so you can display each
member as a point on the historical graph. Then a user can view system perform-
ance trends in a graphical format. Because Armstrong had previously not been col-
lecting performance data, Sue agreed to set up a collection schedule for once a
week to establish some historical data. Bob suggested they use the IBM-supplied
graph formats to show performance graphs (rather than historical graphs), so they
selected option 3 (Display performance data graphs).

Chapter 13. A Problem Analysis Case Study 13-13

 à ð
Display Graphs and Packages

Select one of the following:

1. Display performance data graphs

2. Display historical data graphs

Selection or command
===> 1

F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ
Bob pointed out that the QIBMxxx formats are supplied by IBM. He commented that
a good graph to begin with is the processing unit use versus time (by job type), so
they selected the QIBMCPUTYP member and pressed the Enter key.

à ð
Select Graph Formats and Packages

Library . . . . . . . . . . . QPFRDATA

Type options, press Enter.

1=Select 5=Display sample graph 8=Display package contents

Format/
Option Package Type Description
QIBMPKG PACKAGE IBM GRAPH PACKAGE
QIBMASYNC FORMAT Asynchronous Disk I/O per Second vs. Time
QIBMCMNIOP FORMAT Communications IOP Utilization vs. Time
QIBMCPUPTY FORMAT CPU Utilization vs. Time (Priority)
1 QIBMCPUTYP FORMAT CPU Utilization vs. Time (Job Type)
QIBMDSKARM FORMAT Disk Arm Utilization vs. Time
QIBMDSKIOP FORMAT Disk IOP Utilization vs. Time
QIBMDSKOCC FORMAT Percentage of Disk Occupied vs. Time
QIBMLWSIOP FORMAT Local Workstation IOP Utilization vs. Time
QIBMMFCIOP FORMAT Multifunction IOP (Comm) Util vs. Time
QIBMMFDIOP FORMAT Multifunction IOP (Disk) Util vs. Time
More
F3=Exit F1ð=Restore list F12=Cancel F14=Sort by type F15=Sort by name
F16=Sort by Description

á ñ
On the following display Bob selected the performance data member to be
graphed.

13-14 Performance Tools V4R2

à ð
Select Performance Data Member

Library . . . . . . QPFRDATA

Type options, press Enter.

1=Select

Member
Option Name Description Date Time
1 PERFPROB ð7/ð7/98 14:33:24

Bottom
F3=Exit F5=Refresh F12=Cancel F15=Sort by member
F16=Sort by Description F19=Sort by date/time

á ñ
On the following display Bob pressed F6 (Include all data) and pressed the Enter
key and proceeded to the next display (Figure 13-1 on page 13-16) containing the
graph.

à ð
Select Categories for Performance Graphs

Member . . . . . : PERFPROB
Library . . . . : QPFRDATA

Type options, press Enter. Press F6 to include all data in the graph.
1=Select

Option Category
Job
User ID
Subsystem
Pool
Communications line
Control unit
Functional area

Bottom
F3=Exit F6=Include all data F12=Cancel

á ñ

Chapter 13. A Problem Analysis Case Study 13-15

 CPU Monday’s Performance Data
Utilization CPU Utilization vs. Time (Job Type)
90
80
70
60
50
40
30
20
10
0
08:00 10:00 12:00 14:00 16:00 18:00
Time
Press Enter to continue. Batch
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot System

RV2S075-0

Figure 13-1. CPU Utilization

Here Bob commented to Sue that he would only show the first two and one-half
hours of the collected data to give her a quick idea of what it would look like. He
did this by changing the start and stop parameters to produce the following dis-
plays. The following are the graphs that Bob and Sue elected to look at (that is,
they followed the same previous steps to use the formats QIBMRSP,
QIBMDSKARM, and QIBMDSKIOP).

Interactive
Response Time Interactive Response Time vs Time
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
08:00 08:30 09:00 09:30 10:00 10:30
Time
Press Enter to continue.
F3=Exit F6=Print F9=Overlay Interact
F12=Cancel F16=Save GDF F17=Plot
RV2S076-0

Disk Arm Disk Arm Utilization vs. Time

Utilization
22
20
18
16
14
12
10
8
6
4
2
0
08:00 08:30 09:00 09:30 10:00 10:30
Time
Press Enter to continue.
F3=Exit F6=Print F9=Overlay Maximum
Average
F12=Cancel F16=Save GDF F17=Plot
RV2S080-0

13-16 Performance Tools V4R2

 Disk IOP Disk IOP Utilization vs. Time
Utilization
24
22
20
18
16
14
12
10
8
6
4
2
0
08:00 08:30 09:00 09:30 10:00 10:30
Time
Press Enter to continue.
F3=Exit F6=Print F9=Overlay Maximum
F12=Cancel F16=Save GDF F17=Plot Average

RV2S079-0

Bob then explained the graphs that they had produced and commented to Sue that
the processing unit use, disk arm use, and disk IOP use showed no resource prob-
lems, and that the graphs were a quick way to pick up those types of problems
without having to analyze the reports. The interactive response time graph,
however, did show an abnormality just after 10:00 a.m., which should be investi-
gated further. While the graphs gave a clear overview of how the system was per-
forming, in Armstrong’s situation, more detailed analysis was required of the
gathered data.

Another way to review the collected data was to use the Display Performance Data
(DSPPFRDTA) command. They could quickly see a summary of all the data inter-
actively and isolate data of interest, which they could explore further. Following is
the sequence of steps Bob used to perform further analysis.

à ð
Display Performance Data

Member . . . . . . . . PERFPROB F4 for list

Library . . . . . . QPFRDATA

Elapsed time . . . . : ð9:53:52 Version . . . . . . : 4.ð

System . . . . . . . : SYS4ðð Release . . . . . . : 2.ð
Start date . . . . . : ð7/ð7/98 Model . . . . . . . : 51ð-2144
Start time . . . . . : ð8:ðð:ð1 Serial number . . . : XX-XXXXX
QPFRADJ. . . . . . . : ð QDYNPTYSCD . . . . . : 1

CPU utilization (priority) . . . . . . . . . . . : 23.55

CPU utilization (other) . . . . . . . . . . . . : 14.78
Job count . . . . . . . . . . . . . . . . . . . : 172
Transaction count . . . . . . . . . . . . . . . : 15ð98
Transactions per hour . . . . . . . . . . . . . : 1525
Average response (seconds) . . . . . . . . . . . : 1.45
Disk utilization (percent) . . . . . . . . . . . : 7.29
Disk I/O per second . . . . . . . . . . . . . . : 2ð.9
Logical DB I/O for DDM jobs. . . . . . . . . . . : 52.9

F3=Exit F4=Prompt F5=Refresh F6=Display all jobs F1ð=Command entry

F12=Cancel F24=More keys

á ñ
1. As in the earlier Work with Active Job (WRKACTJOB) and Work with Disk
Status (WRKDSKSTS) commands, Bob found that the overall use of the proc-

Chapter 13. A Problem Analysis Case Study 13-17

 essing unit and disk was not exceptionally high. Also, Sue agreed with Bob that
the average response time of 1.45 seconds was acceptable. (This value
reflects the internal response time of the system and does not include line
transmission time, which is usually not a big difference for local work stations.)
Bob then pressed F15 (Display by interval), which is available after pressing
F24 (More keys).

à ð
Display by Interval

Member . . . . . . . : PERFPROB Elapsed time . . . . : ð9:53:52

Library . . . . . : QPFRDATA

Type options, press Enter. Press F6 to display all jobs.

5=Display jobs

CPU Job Tns Average Disk

Option Date Time Util Count Count Response I/O
_ ð7/ð7/98 ð8:15:ðð 16.11 19 3ð9 .84 486
_ ð7/ð7/98 ð8:29:59 6.97 16 226 .46 2897
_ ð7/ð7/98 ð8:44:57 24.97 25 635 .87 117ð5
_ ð7/ð7/98 ð8:59:56 53.18 28 494 .92 16719
_ ð7/ð7/98 ð9:14:54 62.45 24 318 .7ð 17373
_ ð7/ð7/98 ð9:29:53 71.6ð 31 526 .89 2ð635
_ ð7/ð7/98 ð9:44:51 43.ð6 29 574 .73 9642
_ ð7/ð7/98 ð9:59:49 48.ð8 19 399 .94 94ð9
5 ð7/ð7/98 1ð:14:47 11.97 15 243 4.45 3ð76
_ ð7/ð7/98 1ð:29:45 24.45 23 246 1.49 12556
Bottom
F3=Exit F6=Display all jobs F12=Cancel F13=Display by subsystem
F14=Display by job type

á ñ

The Display by Interval display showed that the system was performing well for
most of the users. Bob quickly rolled through all the displays, searching for the
intervals where the average response times seemed noticeably higher than the
average from the previous Display Performance Data display. Bob explained
that if an intermittent response time problem existed, the shorter sampling
interval should help to highlight it. This logic is not foolproof, he added,
because high transaction counts could still reduce the average response time
and mask a problem.
2. Bob found several intervals where the average response exceeded the 1.45
second average significantly. He reviewed the data to see who was having the
worst response times by:
a. Selecting option 5 (Display jobs) on the Display by Interval display
b. Pressing F24 (More keys)
c. Pressing F21 (Sort by response)

13-18 Performance Tools V4R2

à ð
Display Jobs

Interval . . . . : 1ð:14:47 Member . . . . . : PERFPROB

Elapsed time . . : ð9:53:52 Library . . . . : QPFRDATA

Type options, press Enter.

5=Display job detail

Job CPU Tns Avg Disk

Option Job User Number Type Util Count Rsp I/O
5 DSP18 ORDENTRYð1 ð14273 INT .55 17 15.6 169
5 DSP19 ORDENTRYð2 ð14274 INT 1.55 21 13.4 252
DSP38 CREDITð3 ð14343 INT .71 6 3.ð 389
DSP14 RECVð1 ð14337 INT .ð4 1 2.ð 54
DSP34 SALESð2 ð14339 INT .32 11 1.4 243
DSP41 CREDITð1 ð14285 INT 1.93 24 1.3 493
DSP11 SHIPPINGð1 ð14289 INT .34 8 1.3 251
DSPð1 QSYSOPR ð14276 INT 2.1ð 51 .8 832
DSP22 SALESð1 ð14322 INT .55 28 .7 311
DSP4ð ACTRCVð1 ð14329 INT 2.32 62 .3 216
Bottom
F3=Exit F12=Cancel F15=Sort by job F16=Sort by job type
F19=Sort by CPU F24=More keys

á ñ

DSP18 and DSP19 had very high average responses, but the total number of
disk I/Os for each of these jobs did not appear to be high. Sue confirmed that
these were the order-entry users that had been complaining.
3. Bob entered option 5 (Display job detail) on the Display Jobs display for both
these jobs to further investigate them.

à ð
Display Job Detail

Job . . . . . . . . : DSP18 Job type . . . . . . : INT

User . . . . . . . : ORDENTRYð1 Subsystem . . . . . : QINTER
Number . . . . . . : ð14273 Pool . . . . . . . . : ð4
Member . . . . . . . : PERFPROB Priority . . . . . . : 2ð
Library . . . . . : QPFRDATA Elapsed time . . . . : ð9:53:52

CPU Tns Average Disk Act-> Wait-> Act->

Interval Seconds Count Response I/O Wait Inel Inel
1ð:14:47 2.ð7ð 17 15.6 169 .9 .ð .ð

Bottom
Press Enter to continue.

F3=Exit F11=View 2 F12=Cancel F15=Sort by interval F24=More keys

á ñ

The Display Job Detail display allowed Bob to review the job’s resource
requirements in greater detail. There are actually three views that make up the
total detail picture.

Chapter 13. A Problem Analysis Case Study 13-19

 As Bob scanned the interval data for DSP18, he remarked that the I/O counts
per transaction did not justify the high response time average. Also, the
Wait->Inel and the Act->Inel were both zero, indicating that the job was
obtaining and holding an activity level when needed.
DSP19 showed the same situation.

Bob proceeded to review all the detail information for those two jobs. The following
conclusions were drawn:
Ÿ Both jobs were experiencing extremely wide variations in average response
times.
Ÿ These variations were occurring between 9:30 a.m. and 4:00 p.m. on both of
the days that data was collected.
Ÿ Resource utilizations (processing unit, disk, and main storage) were not exces-
sively high at those times.

Bob mentioned to Sue that these two jobs were definite candidates for further
investigation. The sample data, however, would not give them the detail to deter-
mine the actual cause of the erratic response times. They would need to capture
another type of data using the Start Performance Monitor command. Trace data
would give them greater detail on individual transactions, such as the program that
was most likely running. First, though, they would meet with the order-entry per-
sonnel to get more information on how they use the system and what types of
problems they were experiencing.

Sue made arrangements for all of them to meet.

Review
Bob reviewed the system performance by using the advisor to confirm his earlier
conclusion about system performance. Bob and Sue then used the graphics to
quickly see if there were intervals that showed particularly bad response times and
high use of system resources (higher than the guidelines). This helped Bob and
Sue focus on certain time intervals rather than the whole time period.

Bob decided that another way to quickly review the daily workload and its effect on
system resources was to use the Display Performance Data (DSPPFRDTA)
command. Rather than scan a printed report, he could interactively scan the col-
lected sample data and isolate individual jobs that might be experiencing poor per-
formance.

Normally, sample data could be collected with longer intervals (20 to 30 minutes)
over longer periods (possibly all day) and be used to track the system’s perform-
ance trends. This would enable a company to better manage its system’s resources
and perhaps prevent major performance surprises.

Though Sue informed Bob of the particular jobs to investigate, Bob decided to ini-
tially use the Display Performance Data command to review the overall system sta-
tistics. He then proceeded to focus on individual jobs. Bob could have just as easily
selected only the order-entry jobs to look at. With no previous data to look at, Bob
wanted to view all the jobs to get a feel for Armstrong’s system usage.

Even with this type of data, more detail on what a job is doing must be gathered
using the trace parameter of the Start Performance Monitor command if the cause

13-20 Performance Tools V4R2

 of the problem is to be isolated. Tracing, however, can generate a great deal of
data and could affect the end users when dumping the trace tables. Tracing should
only be used when in problem analysis mode and for shorter periods of time than
when collecting just sample data. Bob wants to talk to the end users to help him
understand the problem and hopefully trace the system at the most opportune time.

Understanding the Symptoms of the Problem

The two order-entry clerks met with Bob and Sue. Bob questioned them on the
types of operations they performed, the problems with the system, and the types of
additions or changes made that may be related to their problems. The following
items were the results of their discussion:
1. Karen and Tim work in the order-entry department as clerks, processing the
orders that are mailed in daily. They have their own private work station area,
which restricts them from viewing each other’s activities. Early in the morning,
the orders are broken into two stacks. Each of the clerks takes a stack and
types it into its own separate transaction file on the system. At the end of the
day, a batch program posts both transaction files to the main order file. Bas-
ically, the orders are not officially in the system until the day after they are
typed.
2. Armstrong gives its largest customers the ability to call in urgent orders.
Normally, only Karen has the authority to take the call and run a menu option
that allows her to enter the order directly into the order file without being first
held in a transaction file. This type of action usually occurs about twelve to
twenty times a day and requires very short interactive response times because
the customer is on the telephone as the order is entered. In the past, Karen
has had no problem with completing an entire telephone order in under 40
seconds.
3. Recently, Armstrong had changed its policy, allowing all of its customers the
ability to call in urgent orders and inquire about order status. This has caused
the number of telephone orders to increase to a point where, now, both Karen
and Tim are authorized to take telephone orders and enter them directly into
the order file. Each of them currently averages 40 calls a day. It seems to them
that the same menu option that took less than a second to bring up the display
can now take 30 to 40 seconds. This caused serious problems with customers
waiting on the telephone.

Bob suggested a plan to help find the cause of the intermittent response time
problem. Because transferring collected trace data to a database file might affect all
of the users on the system, his plan involved controlling both the amount of time
that the monitor ran and when the trace data would be dumped to a file.

Sue would run the Start Performance Monitor (STRPFRMON) command with
tracing active for one hour, and she would have the option to dump the trace table
later. At the end of each run, she would call Karen or Tim and ask if the problem
occurred. If it had, Sue would give Bob a call, and, at the end of the day, explicitly
dump the trace table with the Dump Trace (DMPTRC) command. If the problem
had not occurred, Sue would restart the monitor for another hour (the trace table is
overlaid with the new data). The problem happened often enough, so it should only
take a few attempts to capture the necessary data.

Chapter 13. A Problem Analysis Case Study 13-21

 They all agreed that this would be the best approach to resolve the problem without
affecting the rest of the users. They would start the procedure that afternoon. Bob
made arrangements to be back on Monday morning.

The following shows how Sue entered the Start Performance Monitor
(STRPFRMON) command that afternoon:

à ð
Start Performance Monitor (STRPFRMON)

Type choices, press Enter.

Member . . . . . . . . . . . . . oeproblem Name, \GEN

Library . . . . . . . . . . . . QPFRDATA Name
Text 'description' . . . . . . . Order Entry Problem - Trace On

Time interval (in minutes) . . . 5 5, 1ð, 15, 2ð, 25, 3ð, 35...
Stops data collection . . . . . \ELAPSED \ELAPSED, \TIME, \NOMAX
Days from current day . . . . . ð ð-9
Hour . . . . . . . . . . . . . . 1 ð-999
Minutes . . . . . . . . . . . . ð ð-99
Data type . . . . . . . . . . . \ALL \ALL, \SYS
Select jobs . . . . . . . . . . \ALL \ALL, \ACTIVE
Trace type . . . . . . . . . . . \ALL \NONE, \ALL
Dump the trace . . . . . . . . . \NO \YES, \NO
Job trace interval . . . . . . . .5 .5 - 9.9 seconds
Job types . . . . . . . . . . . \DFT \NONE, \DFT, \ASJ, \BCH...
+ for more values

More...
F3=Exit F4=Prompt F5=Refresh F1ð=Additional parameters F12=Cancel
F13=How to use this display F24=More keys

á ñ
Notice that Trace type was changed to *ALL and Dump the Trace was changed to
*NO.

During the first monitor run that afternoon, the problem did not occur. After the
second monitor run ended, Sue received word that Tim experienced two major
response time problems when trying to enter a telephone order. No more monitor
runs would be needed. At the end of the day (after most of the users had signed
off), Sue issued the Dump Trace (DMPTRC) command to prepare the necessary
data for Bob. The command looked like this:

13-22 Performance Tools V4R2

 à ð
Dump Trace (DMPTRC)

Type choices, press Enter.

Member . . . . . . . . . . . . . oeproblem Name

Library . . . . . . . . . . . . QPFRDATA Name
Job queue . . . . . . . . . . . QCTL Name, \NONE
Library . . . . . . . . . . . QSYS Name, \LIBL, \CURLIB
Text 'description' . . . . . . . \BLANK

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
á ñ
With that accomplished, Sue was ready for Bob’s visit on Monday.

Analyzing the Data—Manager Feature

Bob returned that Monday. After Sue related the activities of Friday afternoon, the
next step was to begin analyzing the data. Because the Display Performance Data
command could only show sample data, Bob chose to print a job summary report
using the Print Transaction Report (PRTTNSRPT) command. To reduce the amount
of printed output to be analyzed, the report was limited to only the order-entry jobs.

à ð
Print Transaction Report (PRTTNSRPT)

Type choices, press Enter.

Member . . . . . . . . . . . . . > oeproblem NAME

Report title . . . . . . . . . . > Order Entry Problem - Trace On

Report type . . . . . . . . . . \SUMMARY \SUMMARY, \TNSACT, \TRSIT...

+ for more values
Time period for report:

Starting time . . . . . . . . \FIRST TIME, \FIRST

Ending time . . . . . . . . . \LAST TIME, \LAST

Additional Parameters

Library . . . . . . . . . . . . QPFRDATA NAME

Report option . . . . . . . . . \SS \SS, \SI, \OZ, \EV, \HV, ' '

\EV
More...
F3=Exit F4=Prompt F5=Refresh F1ð=Additional parameters F12=Cancel
F13=How to use this display F24=More keys

á ñ

Chapter 13. A Problem Analysis Case Study 13-23

 à ð
Print Transaction Report (PRTTNSRPT)

Type choices, press Enter.

Select jobs . . . . . . . . . . \ALL Character value, \ALL

+ for more values

Omit jobs . . . . . . . . . . . > \NONE Character value, \NONE

+ for more values

Select users . . . . . . . . . . ordentry\ Name, generic\, \ALL .

+ for more values

Omit users . . . . . . . . . . . \NONE Name, generic\, \NONE
+ for more values
Select pools . . . . . . . . . . \ALL 1-16, \ALL
+ for more values
Omit pools . . . . . . . . . . . \NONE 1-16, \NONE
+ for more values
Select functional areas . . . . \ALL

+ for more values

Omit functional areas . . . . . \NONE

More...
F3=Exit F4=Prompt F5=Refresh F1ð=Additional parameters F12=Cancel
F13=How to use this display F24=More keys

á ñ
The following pages show selected sections from the Print Transaction Report
output with areas that Bob highlighted and discussed with Sue.

Job Summary
The job summary for the order-entry department had the report and results shown
in Figure 13-2 on page 13-29.

The following information was extracted from this report:

Ÿ The average response time for ORDENTRY01 was 3.2 seconds and at least
one of its transactions lasted 38.2 seconds. ORDENTRY01 is Tim’s user
profile. What was happening to that job?
Ÿ The average processing unit time per transaction was .24 seconds with at least
one transaction using .42 processor seconds. These times could not be the
reason for the poor performance.
Ÿ ORDENTRY01 showed an average of 15 disk I/O operations per transaction
and a worst case of 51 disk I/Os per transaction. Using .05 seconds as an
average disk I/O service time, these numbers do not justify the exceptionally
long response times.

Sue asked Bob about the two in the Lock Conflict column. Bob remarked that this
value indicated the number of times that ORDENTRY01 needed to wait for an
object being held by another job.

This first page of the report indicated to Bob that Tim was definitely experiencing
poor response times, especially because his workload was similar to Karen’s. Bob
needed further information on what components of the response time were causing
the problem.

13-24 Performance Tools V4R2

System Summary Data
Bob scanned further down the report to look at interactive transaction averages and
exceptional-wait breakdown by job type. See Figure 13-3 on page 13-29.

The following performance information was extracted from this report:

Ÿ The average response time for 57 transactions was 1.613 seconds. This does
not appear to be too high.
Ÿ However, the amount of processing unit time and disk time per transaction do
not justify the 1.613 response time.
Ÿ Of the 1.613 seconds, 1.314 is spent in what is known as exceptional wait.
The Excp Wait/Tns time is that portion of response time that cannot be attri-
buted to processing unit or disk usage and is caused by contention for internal
system resources (for example, waiting for a message queue). Normally, this
value should be less than 10 percent of the total average response time.
Ÿ Almost all of the exceptional wait time is being spent in the Lock Wait category.
(Remember Sue’s question?)

Bob saw further data supporting the existence of a problem. He explained to Sue
that these high numbers still reflected averages.

Analysis by Interactive Response Time

The next section Bob looked at in the report (Analysis by Interactive Response
Time) would help define the makeup of the transactions as shown in Figure 13-5
on page 13-29.

The following information regarding response time was extracted from this trace
report:
Ÿ Of the 57 transactions measured, only two were greater than 10 seconds, and
together they averaged 36.664 seconds.
Ÿ Almost all of that time (36.497 seconds) was spent as exceptional wait time.
Remember that exceptional wait time is nonproductive time. What were those
two transactions doing?

This section of the report allows Bob to help evaluate performance versus objec-
tives. Bob sees that both jobs are actually getting excellent service most of the
time. Two transactions, however, seem to be the source of the high averages.

Individual Transaction Statistics

Bob needed to find out more about those two transactions, so he scanned further
down the report to the Individual Transaction Statistics section as shown in
Figure 13-4 on page 13-29.

This section lists the individual transactions of various statistics (longest response
time, processing unit, service time, and so on).
Ÿ ORDENTRY01 had two very long response times during the collection period,
one at 14:23:27 (38.157 seconds) and the other at 14:32:08 (35.171 seconds).
Ÿ Bob noticed that, at both of these times, the program involved in the transaction
was ORD110.

Chapter 13. A Problem Analysis Case Study 13-25

 Transactions with Longest Lock Wait Time
Bob then looked at the Transactions with the longest lock wait time.

At the same time as those long transactions, ORDENTRY01 experienced extremely

long lock waits. In fact, almost the entire time spent in the transactions was spent
waiting on locks. Again, program ORD110 was involved.

Bob and Sue now had an idea of what was causing the problem. But what kind of
lock was it and why couldn’t ORDENTRY01 get that lock? More questions needed
answering.

Longest Seize/Lock Conflicts

Bob’s next step was to go to the Longest Seize/Lock Conflicts section of the Job
Summary Report. An example of this section is in Figure 13-7 on page 13-30.

This section lists the longest seize/lock conflicts in descending order with the time it
happened, the requesting job, the holding job, and the held object.
Ÿ The two transactions with the long response times for ORDENTRY01 are listed
here as the two longest instances of a lock conflict. The times coincide with
those earlier in the report.
Ÿ The holding job (preventing ORDENTRY01 from obtaining the necessary lock)
in both instances was ORDENTRY02 (Karen’s interactive job).
Ÿ The lock request is for a file called ORDCTL in library OELIB.

Bob narrowed the problem to a conflict between the two jobs ORDENTRY01 and
ORDENTRY02. However, Bob wanted to get a little more information on the trans-
actions that both ORDENTRY01 and ORDENTRY02 were running during the lock
conditions. Further detail on the transactions in question could be explored by
running another Print Transaction Report, this time asking for transition detail infor-
mation. This report normally produces a great deal of output. The report could be
efficiently reviewed by selecting only the jobs and times involved with the problem.

Bob entered the Print Transaction Report (PRTTNSRPT) command to get the fol-
lowing display:

13-26 Performance Tools V4R2

à ð
Print Transaction Report (PRTTNSRPT)

Type choices, press Enter.

Member . . . . . . . . . . . . . > OEPROBLEM Name

Report title . . . . . . . . . . > 'Order Entry Problem - Transitional Report'

Report type . . . . . . . . . . \TRSIT \SUMMARY, \TNSACT, \TRSIT...

+ for more values
Time period for report:

Starting time . . . . . . . . 142ððð TIME, \FIRST

Ending time . . . . . . . . . 1435ðð TIME, \LAST

Additional Parameters

Library . . . . . . . . . . . . QPFRDATA Name

Report option . . . . . . . . . \SS \SS, \SI, \OZ, \EV, \HV, ' '

+ for more values

More...
F3=Exit F4=Prompt F5=Refresh F1ð=Additional parameters F12=Cancel
F13=How to use this display F24=More keys

á ñ

à ð
Print Transaction Report (PRTTNSRPT)

Type choices, press Enter.

Select jobs . . . . . . . . . . \ALL Character value, \ALL

+ for more values

Omit jobs . . . . . . . . . . . > \NONE Character value, \NONE

+ for more values

Select users . . . . . . . . . . ordentry\ Name, generic\, \ALL .

+ for more values

Omit users . . . . . . . . . . . \NONE Name, generic\ \NONE
+ for more values
Select pools . . . . . . . . . . \ALL 1-16, \ALL
+ for more values
Omit pools . . . . . . . . . . . \NONE 1-16, \NONE
+ for more values
Select functional areas . . . . \ALL

+ for more values

Omit functional areas . . . . . \NONE

Bottom
F3=Exit F4=Prompt F5=Refresh F1ð=Additional parameters F12=Cancel
F13=How to use this display F24=More keys

á ñ
Notice that this time the output had been reduced to only showing information
about the ORDENTRY01 and ORDENTRY02 jobs between 14:20:00 and 14:35:00.

Chapter 13. A Problem Analysis Case Study 13-27

 Transition Report for ORDENTRY01
The example report in Figure 13-8 on page 13-31 shows sections of the Transition
Report resulting from the PRTTNSRPT command just issued.

First, Bob scanned the Transition Report for ORDENTRY01 and noticed:
Ÿ At 14.23.28.135, ORDENTRY01 went into a lock wait for 37.819 seconds
because of a request for file ORDCTL, over which job ORDENTRY02 had a
lock.
Ÿ ORD110 appeared to be the order-entry program asking for the file. Programs
starting with the letter Q (for example, QDBGETSQ) are normally IBM-supplied
system service routines.
Ÿ The same condition appeared to be happening at 14.32.08.691.

Transition Report for ORDENTRY02

Next, Bob scanned the Transition Report for ORDENTRY02 and noted that at the
times surrounding the lock waits for ORDENTRY01 (14.23.17.455 to 14.24.05.954
and 14.31.48.059 to 14.32.43.665), ORDENTRY02 was also running ORD110. An
example of this report is in Figure 13-9 on page 13-32.

Bob felt that exploring program ORD110 might help them understand why the locks
were occurring. Sue took Bob to the data processing department to talk with
Armstrong’s lead programmer.

Review
Bob and Sue together determined that they would first select only the order-entry
jobs when producing the transaction summary reports. They could do this only
because they had a good idea of the jobs in question. Under different circum-
stances, using all of the jobs as input to the report may be necessary. Limit the
number of transactions to analyze whenever possible.

Through the different sections of the report, Bob was able to isolate not only the job
in trouble, but also the individual transactions, times, and programs involved in the
problem.

It is important to note that Bob did not stop at finding the job having a problem. It is
much more critical that the cause of the problem be found. ORDENTRY01 is the
job preventing ORDENTRY02 from obtaining service. By looking at job transition
information and matching times, the suspected program (ORD110) was identified.

With this information, Bob and Sue could now approach the application developers
for a solution.

Case Study Data Reports—Manager Feature

Analyzing the data in the following example reports helped Bob and Sue under-
stand their problem.

13-28 Performance Tools V4R2

 Job Summary Reports
Job Summary Report ð7/ð7/98 1ð:33:42
Job Summary Page ððð1
Order-Entry Problem - Trace On

Member . . . : OEPROBLEM Model/Serial . . : 51ð-2144/XX-XXXXX Main storage . . : 384.ð M Started . . . . : ð5/13/98 14:ð3:19
Library . . : QPFRDATA System name . . . : SYS4ðð Version/Release : 4/ 2.ð Stopped . . . . : ð5/13/98 14:57:5ð

\On/Off\ T P P Tot Response Sec CPU Sec ---- Average DIO/Transaction ----- Number K/T
Job User Job y t r Nbr ------------- ------------------ ------ Synchronous ----- --Async-- Cft /Tns
Name Name Number Pl p y g Tns Avg Max Util Avg Max DBR NDBR Wrt Sum Max Sum Max Lck Sze Sec
---------- ---------- ------ -- -- -- - ---- ------ ------ ---- ------ ------ ---- ---- ---- ---- ---- ---- ---- ------ -----
DSP18 ORDENTRYð1 ð31288 ð2 I 2ð N 26 3.2 38.2 .4 .24 .42 4 8 3 15 51 1 13 2 2 85
DSP19 ORDENTRYð2 ð31289 ð2 I 2ð N 31 .3 2.ð .5 .26 .87 3 9 2 14 29 74

Figure 13-2. Job Summary: Order-Entry Problem - Trace On

DATA FOR SELECTED TIME INTERVAL (OR TOTAL TRACE PERIOD IF NO TIME SELECTION).

INTERACTIVE TRANSACTION AVERAGES BY JOB TYPE.

T Avg CPU/ ----- Sync Disk I/O Rqs/Tns ----- Async W-I Excp Key/ Active Est
y Nbr Nbr Pct Tns Rsp Tns DB DB NDB NDB DIO Wait Wait Think K/T Of
p Prg Jobs Tns Tns /Hour (Sec) (Sec) Read Write Read Write Sum /Tns /Tns /Tns /Tns /Tns AWS
-- --- ---- ------ ----- ----- ------ ------ ----- ----- ----- ----- ----- ----- ------ ------- ------ ------ ---
I NO 2 57 1ðð.ð 62 1.613 .253 3 2 8 1 14 ð .ððð 1.314 79.ð92 55.254

EXCEPTIONAL WAIT BREAKDOWN BY JOB TYPE.

A-I Short Short Seize Lock Event Excs EM327ð DDM Svr Other
Wait Wait WaitX Wait Wait Wait ACTM Wait Wait Wait
Type Purge /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns /Tns
---- ----- ------ ------ ------ ------ ------ ------ ------ ------ ------- ------
I NO .ððð .ð33 .ððð .ðð3 1.277 .ððð .ðð1 .ððð .ððð .

Figure 13-3. Data for Selected Time Interval

Job Summary Report ð7/ð7/98 18:33:42

Individual Transaction Statistics Page ðð21
Order-Entry Problem - Trace On

Member . . . : OEPROBLEM Model/Serial . . : 51ð-2144/XX-XXXXX Main storage . . : 384.ð M Started . . . . : ð5/13/98 14:ð3:19
Library . . : QPFRDATA System name . . . : SYS4ðð Version/Release : 4/ 2.ð Stopped . . . . : ð5/13/98 14:57:5ð

TRANSACTIONS WITH LONGEST RESPONSE TIMES

Rank Value Time Program Job Name User Name Number Pool Type Priority
---- -------- ------------ ---------- ---------- ---------- ------ ---- ---- ---------
1 38.157 14.23.27.921 ORD11ð DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
2 35.171 14.32.ð8.618 ORD11ð DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
3 2.274 14.36.11.625 QUIINMGR DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
4 1.951 14.41.22.7ð5 QUIINMGR DSP19 ORDENTRYð2 ð31289 ð2 I 2ð
5 1.543 14.ð5.56.163 QUIINMGR DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
6 1.ð41 14.ð5.47.886 QUIINMGR DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
7 .777 14.35.55.734 QUIINMGR DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
8 .567 14.33.ð8.82ð QUIINMGR DSP19 ORDENTRYð2 ð31289 ð2 I 2ð
9 .562 14.35.4ð.131 QUIINMGR DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
1ð .491 14.29.15.ð71 QUIINMGR DSP19 ORDENTRYð2 ð31289 ð2 I 2ð

Figure 13-4. Individual Transaction Statistics

ANALYSIS BY INTERACTIVE RESPONSE TIME.

Avg Cum Avg Cum ----- Sync Disk I/O Rqs/Tns ----- Async Excp Avg
Rsp Nbr Pct Pct CPU CPU CPU DB DB NDB NDB DIO Wait K/T
Category /Tns Tns Tns Tns /Tns Util Util Read Write Read Write Sum /Tns /Tns /Tns
---------------- ------- ----- ----- ----- ------- ----- ----- ----- ----- ----- ----- ----- ----- ------- --------
Sub-Second .332 51 89.5 89.5 .229 .2 .2 1 1 2 4 .ð37 51.979
1 - 1.999 Sec 1.512 3 5.3 94.8 .498 .2 7 12 9 28 1 1ð.ð28
2 - 2.999 Sec 2.274 1 1.8 96.6 .419 .2 2 25 1 23 51 13 1.185
3 - 4.999 Sec 96.6 .2
5 - 9.999 Sec 96.6 .2
GE 1ð Seconds 36.664 2 3.5 1ðð.1 .ð91 .2 2 2 36.497 233

Figure 13-5. Analysis by Interactive Response Time

Chapter 13. A Problem Analysis Case Study 13-29

TRANSACTIONS WITH LONGEST LOCK WAIT TIME

Rank Value Time Program Job Name User Name Number Pool Type Priority
---- -------- ------------ ---------- ---------- ---------- ------ ---- ---- ---------
1 37.822 14.23.27.921 ORD11ð DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
2 34.977 14.32.ð8.618 ORD11ð DSP18 ORDENTRYð1 ð31288 ð2 I 2ð
3
4
5
6
7
8
9
1ð

Figure 13-6. Transactions with Longest Wait Time

Job Summary Report ð7/ð7/98 1ð:33:42

Longest Seize/Lock Conflicts Page ðð26
Order-Entry Problem - Trace On

Member . . . : OEPROBLEM Model/Serial . . : 51ð-2144/XX-XXXXX Main storage . . : 384.ð M Started . . . . : ð5/13/98 14:ð3:19
Library . . : QPFRDATA System name . . . : SYS4ðð Version/Release : 4/ 2.ð Stopped . . . . : ð5/13/98 14:57:5ð

Job User Job Holder- Job Name.. User Name. Number Pool Type Pty
Rank Value Time Name Name Number Pl Typ Pty S/L Object- Type.. Library... File...... Member.... RRN......
---- -------- ------------ ---------- ---------- ------ -- --- --- --- ---------------------------------------------------------
1 37.819 14.23.28.135 DSP18 ORDENTRYð1 ð31288 ð2 I 2ð L HOLDER- DSP19 ORDENTRYð2 ð31289 ð2 I 2
OBJECT- DS OELIB ORDCTL ORDCTL
2 34.974 14.32.ð8.691 DSP18 ORDENTRYð1 ð31288 ð2 I 2ð L HOLDER- DSP19 ORDENTRYð2 ð31289 ð2 I 2ð
OBJECT- DS OELIB ORDCTL ORDCTL
3 .ð9ð 14.32.43.67ð DSP18 ORDENTRYð1 ð31288 ð2 I 2ð S HOLDER- DSP19 ORDENTRYð2 ð31289 ð2 I 2ð
OBJECT- DS OELIB ORDCTL ORDCTL
4 .ð89 14.24.ð5.959 DSP18 ORDENTRYð1 ð31288 ð2 I 2ð S HOLDER- DSP19 ORDENTRYð2 ð31289 ð2 I 2ð
OBJECT- DS OELIB ORDCTL ORDCTL

Figure 13-7. Longest Seize/Lock Conflicts

Transition Reports

13-30 Performance Tools V4R2

 Transition Report ð7/ð7/98 11:41:45
Order-Entry Problem - Trace On Page ððð1

Member . . . : OEPROBLEM Model/Serial . . : 51ð-2144/XX-XXXXX Main storage . . : 384.ð M Started . . . . : ð5/13/98 14:2ð:ðð
Library . . : QPFRDATA System name . . . : SYS4ðð Version/Release : 4/ 2.ð Stopped . . . . : ð5/13/98 14:35:ðð
Job name . . : DSP18 User name . . . . : ORDENTRYð1 Job number . . . : ð31288 TDE/Pl/Pty/Prg . : ð1B4 ð2 2ð

Job type . . : I Elapsed Time -- Seconds Sync/Async Phy I/O -MPL-

----------------------- ------------------------ C I Last 4 Programs in Invocation Stack
State Wait Long Active Inel CPU DB DB NDB NDB u n ------------------------------------------
Time W A I Code Wait /Rsp\ Wait Sec Read Wrt Read Wrt Tot r l Last Second Third Fourth
------------ ----- ---- ------- ------- ------- -------- ---- ---- ---- ---- ---- -- -- ---------- ---------- --------------------

\\\\\ NO Is The New Purge Attribute. -

14.23.27.921 ->A 1
14.23.28.135 LKWT \/ 37.819/\ HOLDER-- DSP19 ORDENTRYð2 ð31289
OBJECT- DS OELIB ORDCTL ORDCTL
14.23.28.136 W<- LKW .215 .ð78 1 QDBGETSQ ORD11ð QUIMNDRV QUICMENU
14.24.ð5.957 ->A 37.822 2
14.24.ð5.959 SZWT \/ .ð89/\ HOLDER-- DSP19 ORDENTRYð2 ð31289
OBJECT- DS OELIB ORDCTL ORDCTL
14.24.ð6.ð77 W<- .12ð .ð14 2 QT3REQIO QWSGET ORD11ð QUMNDRV
---------- ORD11ð 38.157\ .ð92 ð ð 4 ð 4\

<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< Later in the report... >>>>>>>>>>>

;>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

14.32.ð8.618 ->A 167.297 1

14.32.ð8.691 LKWT \/ 34.974/\ HOLDER-- DSP19 ORDENTRYð2 ð31289
OBJECT- DS OELIB ORDCTL ORDCTL
14.32.ð8.692 W<- LKW .ð75 .ð75 1 QDBGETSQ ORD11ð QUIMNDRV QUICMENU
14.32.43.669 ->A 34.977 2
14.32.43.67ð SZWT \/ .ð9ð/\ HOLDER-- DSP19 ORDENTRYð2 ð31289
OBJECT- DS OELIB ORDCTL ORDCTL
14.32.43.788 W<- .119 .ð14 2 QT3REQIO QWSGET ORD11ð QUIMNDRV
---------- ORD11ð 35.171\ .ð89 ð ð ð ð ð\
14.33.24.8ð6 ->A 41.ð18 .ðð1 .ðð1 1
2
14.33.24.918 W .112 .ð19 1 QT3REQIO QWSCLOSE QDMCLOSE ORD11ð
PAG= ð XSum= ð PWrt= 2
14.33.24.949 A .ð3ð .ðð1 1
14.33.25.ð12 W .ð64 .ð57 1 QT3REQIO QWSPUT QUIMNDRV QUICMENU
14.33.25.ð72 A .ð59 .ðð1 1
14.33.25.ð87 W<- .ð15 .ð14 1 QT3REQIO QWSGET QUIMNDRV QUICMENU
---------- QUIMNDRV .281\ .ð93 ð 2 ð ð 2\ PAG= ð XSum= ð PWrt= 2

Figure 13-8. Transition Report for ORDENTRY01

Chapter 13. A Problem Analysis Case Study 13-31

 Transition Report ð7/ð7/98 11:31:45
Order-Entry Problem - Trace On Page ððð3

Member . . . : OEPROBLEM Model/Serial . . : 51ð-2144/XX-XXXXX Main storage . . : 384.ð M Started . . . . : ð5/13/98 14:2ð:ðð
Library . . : QPFRDATA System name . . . : SYS4ðð Version/Release : 4/ 2.ð Stopped . . . . : ð5/13/98 14:35:ðð
Job name . . : DSP19 User name . . . . : ORDENTRYð2 Job number . . . : ð31289 TDE/Pl/Pty/Prg . : ð1C3 ð2 2ð

Job type . . : I Elapsed Time -- Seconds Sync/Async Phy I/O -MPL-

----------------------- ------------------------ C I Last 4 Programs in Invocation Stack
State Wait Long Active Inel CPU DB DB NDB NDB u n ------------------------------------------
Time W A I Code Wait /Rsp\ Wait Sec Read Wrt Read Wrt Tot r l Last Second Third Fourth
------------ ----- ---- ------- ------- ------- -------- ---- ---- ---- ---- ---- -- -- ---------- ---------- --------------------

\\\\\ NO Is The New Purge Attribute. -

14.22.44.ð88 A .ð27 .ðð1 1

14.22.44.145 W .ð57 .ð55 1 QT3REQIO QWSPUT QUIMNDRV QUICMENU
14.22.44.222 A .ð77 1
14.22.44.237 W<- .ð16 .ð14 1 QT3REQIO QWSGET QUIMNDRV QUICMENU
---------- QUIMNDRV .35ð\ .ð88 ð 2 ð 1 3\ PAG= ð XSum= ð PWrt= 3
14.23.17.455 ->A 33.217 .ðð1 .ðð1 1
14.23.17.538 W<- .ð83 .ð82 1 QT3REQIO QWSGET ORD11ð QUIMNDRV
---------- ORD11ð .ð84\ .ð83 ð ð ð ð ð\
14.24.ð5.885 ->A 48.347 1
14.24.ð5.954 LKRL WAITER- DSP18 ORDENTRYð1 ð31288 37.819 SECS
OBJECT- DS OELIB ORDCTL ORDCTL
14.24.ð6.ð49 SZRL WAITER- DSP18 ORDENTRYð1 ð31288 .ð89 SECS
OBJECT- DS OELIB ORDCTL ORDCTL
14.24.ð6.ð61 W .176 .ð18 2 QT3REQIO QWSCLOSE QDMCLOSE ORD11ð
PAG= ð XSum= ð PWrt= 3
14.24.ð6.ð94 A .ð33 .ðð1 1
14.24.ð6.151 W .ð57 .ð55 1 QT3REQIO QWSPUT QUIMNDRV QUICMENU
14.24.ð6.228 A .ð77 1
14.24.ð6.243 W<- .ð16 .ð15 1 QT3REQIO QWSGET QUIMNDRV QUICMENU
---------- QUIMNDRV .359\ .ð89 ð 2 ð 1 3\ PAG= ð XSum= ð PWrt= 3

<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< Later in the report... >>>>>>>>>>>

;>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

---------- QUIMNDRV .219\ .135 ð ð ð ð ð\

14.31.48.ð59 ->A 148.91ð 1
14.31.48.138 W<- .ð79 .ð77 1 QT3REQIO QWSGET ORD11ð QUIMNDRV
---------- ORD11ð .ð79\ .ð77 ð ð ð ð ð\
14.32.43.598 ->A 55.46ð .ðð1 .ðð1 1
14.32.43.665 LKRL WAITER- DSP18 ORDENTRYð1 ð31288 34.974 SECS
OBJECT- DS OELIB ORDCTL ORDCTL
14.32.43.76ð SZRL WAITER- DSP18 ORDENTRYð1 ð31288 .ð9ð SECS
OBJECT- DS OELIB ORDCTL ORDCTL
2 1
14.32.43.773 W .175 .ð18 2 QT3REQIO QWSCLOSE QDMCLOSE ORD11ð
PAG= ð XSum= ð PWrt= 3
14.32.43.8ð6 A .ð34 .ðð1 1
14.32.43.865 W .ð58 .ð57 1 QT3REQIO QWSPUT QUIMNDRV QUICMENU
14.32.43.942 A .ð77 1
14.32.43.957 W<- .ð15 .ð14 1 QT3REQIO QWSGET QUIMNDRV QUICMENU
---------- QUIMNDRV .36ð\ .ð91 ð 2 ð 1 3\ PAG= ð XSum= ð PWrt= 3

Figure 13-9. Transition Report for ORDENTRY02

Finding the Cause and Correcting the Problem

Mike Brown was the senior applications programmer for Armstrong. Although not
directly responsible for the order-entry application code, he could get the necessary
documentation on ORD110 for Bob so that the record lock problem could be ana-
lyzed.
1. Mike and Bob reviewed the program’s flowchart and source code and were
able to determine that ORD110 is an RPG/400 program that opens four files:
CUSMSTL Customer master file, input only
ORD110D Work station display file, input and output
ORDFILL Open orders file, output only (add)

13-32 Performance Tools V4R2

 ORDCTL Order control file, update
2. When Enter a new order is selected, ORD110 gets the single control record
from ORDCTL, which contains the next order number. Every order must have a
unique order number.
3. The order-entry clerk responds to a prompt from display file ORD110D, asking
the clerk for the customer number. This customer number is then used by the
program to get customer information from the CUSMSTL logical file, which in
turn is presented to the clerk.
4. The clerk enters the necessary order data. When finished, the data is added to
ORDFILL as a new order.
5. Finally, the order number field of the control record is incremented by one and
written back to ORDCTL. This allows the next order entered to have the next
higher order number.

To Bob and Mike, the record lock conflict for ORDCTL was very obvious. With only
one clerk using ORD110, the lock on the control record for update did not present
any problem. Armstrong’s original policy of having Karen as the only authorized
user of ORD110 ensured that only one clerk would use ORD110. The other orders
received through the mail would not be assigned an order number until the night
time batch job.

With the change in policy allowing multiple clerks to access ORD110, two clerks
could now attempt to enter an order at the same time. Only one clerk, though,
could have the ORD110D display available to them because they would first need
an exclusive lock on the control record. This record would be locked for the entire
order process. The requesting job’s display would be inhibited while the holding job
completed its order. Because the process only lasted about 30 seconds, the control
record was released before another requesting job timed out (the default wait time
on a record lock is 60 seconds). Had the time-out occurred, a function check would
have alerted the data processing department to a lock problem much sooner.

Mike quickly created a coding correction for ORD110 such that the reading, incre-
menting, and updating of the control record would be done at the end of the order
process. This would allow the records to be locked and released in an instant and
allow other jobs to do the same. Later on, a more efficient technique, such as using
a data area to store the control information, could be further explored.

Bob suggested to Sue that she run the performance monitor again for the next day
to measure the results of the change. The tracing option would be off and the
monitor should run for the whole day. The order-entry department was to notify her
if the response time situation occurred again. At a later date, Bob would return to
work with Sue on developing some system monitoring practices that Armstrong
should use with Performance Tools.

Final Review
The case study you just read is an example of one person’s approach to solving a
typical application performance problem. The methodology was based on several
logical steps:
1. Understand the symptoms of the problem
Initially, Bob was made aware of a problem with very little information to help
him to solve it. His first actions involved using commands to determine how well

Chapter 13. A Problem Analysis Case Study 13-33

 the system was reacting to the overall workload. By isolating the users having
problems and talking to them, he was able to identify their objectives and sub-
stantiate the existence of a problem. The information he collected through the
interview with the order-entry personnel was critical in effectively analyzing the
situation.
2. Use tools to measure and define the problem
Performance Tools proved instrumental in determining not only what jobs were
part of the problem, but also what programs were involved and at what times.
Problems like poor response time have a definite cause and, in most cases, the
available system tools can help capture and report the vital information.
Selecting specific times and jobs enabled Bob to reduce the amount of data
that had to be analyzed.
3. Isolate the cause and correct the problem
Bob and Mike carefully analyzed the problem and examined the application and
database design to develop a solution. They also ensured that the solution did
not produce negative effects for other jobs or cause incorrect data in the busi-
ness operations.
4. Use tools to verify the problem is corrected
As mentioned earlier, Sue was to run the Start Performance Monitor
(STRPFRMON) command the next day to measure the results of the change. If
new problems appeared, the above steps would be repeated until the solution
became acceptable.

Armstrong’s story is an example of a single, isolated problem. In some cases, a

system may have many different problems occurring at the same time. Prioritize
the problems to select which items to investigate first. When those problems are
resolved, go after the next in line until the situation no longer justifies the time and
effort.

Another situation may be that a big problem is the result of an accumulation of

many little design flaws. Some poor programming techniques may not affect one
user much, but if multiplied by many jobs running at the same time, the result can
be dramatic.

Finally, the fact may be that the resources are seriously overcommitted and that it
is time for a model upgrade or another disk controller. Use the capacity planning
option of Performance Tools to help you determine the additional resources needed
to meet the performance objectives.

Learn the proper usage of the tools available to you, and start to put into place a
strategy that will help you get the most out of your AS/400 system.

13-34 Performance Tools V4R2

Chapter 14. Working with Historical Data—Agent Feature
This chapter describes the commands used to maintain historical data. Historical
data is a summary of the performance data created by the performance monitor.

The Agent feature allows you to create historical data, which may then be analyzed
on another AS/400 system that has the Manager feature installed. You can also
analyze the data using your own programs or queries. Appendix E, Managing
AS/400 System Performance in a Network, provides a customer scenario which
illustrates how a customer can create historical data at remote AS/400 sites. The
historical data is then transmitted to a central AS/400 system for analysis using the
Manager feature. The Manager feature provides the capability to present historical
data in the form of graphics. Appendix D, Comparison of Performance Tools pro-
vides more information on the functions provided in Performance Tools.
Note: Files are created to contain the historical data. For each performance
member with historical data, there is a single value for each type of informa-
tion that can be graphed for each day of the member’s performance col-
lection period. Thus, the amount of data is reduced and summarized into
the historical files. Once you have historical data for a member, you may
choose to delete the performance data (DLTPFRDTA) created through the
initial performance data collection to free file storage space.

Since historical data can help show trends in your system’s performance, it is
recommended that you create historical data in a given library for members that are
collected at the same time. (For example, you might want to compare data that was
all collected on Wednesdays from 8:00 a.m. to 12:00 p.m., whereas you probably
would not want a historical data with one member collected on Wednesday from
8:00 a.m. to 12:00 p.m. and the other on Saturday from 1:00 to 5:00 p.m.)

To simplify data management, consider using separate libraries for comparable col-
lections of data.

If you select option 3 (Work with historical data) on the Performance Tools
Graphics menu, the Work with Historical Data display appears.

 Copyright IBM Corp. 1998 14-1

 à ð
Work with Historical Data

Library . . . . . . QPFRDATA

Type options, press Enter.

1=Create historical data 4=Delete historical data

Member Historical
Option Name Data Date Time
_ Q95318ð843 NO 11/14/95 ð8:43:15
_ Q953171ð5ð NO 11/13/95 1ð:51:ðð
_ SATDATA YES 11/11/95 1ð:42:48
_ TESTDATA YES 11/11/95 1ð:26:12
_ NOV111995 NO 11/11/95 ð9:57:27
_ Q95315ð955 NO 11/11/95 ð9:55:41
_ FRIDAY YES 11/1ð/95 11:17:ð3
_ Q953132332 YES 11/ð9/95 23:32:19
_ Q9531214ð7 YES 11/ð8/95 14:ð7:11
_ Q953121142 NO 11/ð8/95 11:42:3ð
_ Q953111538 NO 11/ð7/95 15:39:ð2
More...
F3=Exit F5=Refresh F11=Display text F12=Cancel F15=Sort by Member
F16=Sort by text

á ñ
The member name, a historical data indicator, and the date and time you collected
each set of performance data appear on this display. To display the member text
description, press F11 (Display text). If you cannot find the data you want to work
with, use the appropriate function key to sort the sets of performance and historical
data. You can sort them by member name, text description, or by the date and time
the member was created. When you find the data you want to work with, indicate
the function you want to perform by typing the appropriate option.

If you are searching for performance or historical data located in a library that is
different from the one currently listed in the Library field at the top of the display,
type a new library name in the Library field and press the Enter key. A list of per-
formance and historical data members available in the library you specified
appears. You can then select one of them to work with.
Note: All of the members in the historical data must have unique names. If you
create a member that has the same name as a historical data member, you
may want to change the name by using the Copy Performance Data
(CPYPFRDTA) command to use the new member for historical purposes.

It is best to use the created name option (*GEN) on the STRPFRMON command to
make sure that the names of your performance data members are unique.

Create Historical Data

To create historical data for performance members, type a 1 (Create) by the
members, and press the Enter key. The Confirm Create of Historical Data display
appears.

14-2 Performance Tools V4R2

 à ð
Confirm Create of Historical Data

Library . . . . . : QPFRDATA

Press Enter to confirm your choices for 1=Create.

Press F12=Cancel to return to change your choices.

Member Historical
Option Name data Date Time
1 Q95318ð843 NO 11/14/95 ð8:43:15
1 Q953171ð5ð NO 11/13/95 1ð:51:ðð
1 SATDATA YES 11/11/95 1ð:42:48

Bottom
F11=Display text F12=Cancel

á ñ
On this display, press the Enter key to create historical data for the members. Once
historical data has been created for a member, you can delete the original perform-
ance data using the Delete Performance Data (DLTPFRDTA) command if the data
is not needed for performance analysis, capacity planning, or performance
graphing.

Delete Historical Data

To delete the historical data created by the Create Historical Data command, type a
4 (Delete) by members that contain historical data, and press the Enter key. This
does not delete the original performance data.
Note: If the performance data for a member no longer exists, you cannot re-create
historical data for that member after the historical data has been deleted.

Chapter 14. Working with Historical Data—Agent Feature 14-3

14-4 Performance Tools V4R2
Chapter 15. Managing the Performance Data—Agent Feature
If you choose the Manage performance data option on the IBM Performance Tools
menu for the Agent feature, the Manage Performance Data display appears.

à ð
Manage Performance Data

Select one of the following:

1. Delete performance data

2. Copy performance data
3. Convert performance data

Selection or command
===> _____________________________________________________________________
__________________________________________________________________________
F3=Exit F4=Prompt F9=Retrieve F12=Cancel

á ñ
From this display you can manage the performance data collected by the perform-
ance monitor.

Delete Performance Data

Use option 1 (Delete performance data) on the Manage Performance Data display
to delete performance data that you no longer need on your system. When you
choose option 1, the Delete Performance Data display appears.

 Copyright IBM Corp. 1998 15-1

 à ð
Delete Performance Data

Library . . . . . . QPFRDATA__

Type option, press Enter.

4=Delete
Option Member Text Date Time
_ XYZ 12/15/95 14:ð5:55
_ PERFTESTC4 2 hours w/ 5 minute intervals 12/15/95 ð8:ð5:48
_ PERFTESTC3 Duration of 2 hours 12/14/95 ð9:21:44
_ PERFTESTC2 12/11/95 14:42:46

Bottom
F3=Exit F5=Refresh F12=Cancel F15=Sort by member F16=Sort by text

á ñ
The members that appear on this display are those used on the Start Performance
Monitor (STRPFRMON) command for the keyword MBR when data was collected.
To delete a member from this list, type a 4 (Delete) next to the appropriate member
and press the Enter key. The member you delete is deleted from the following data
collection files:

QAPGSUMD QAPMHDLC QAPMSTND

QAPMAPPN QAPMHDWR QAPMSTNE
QAPMASYN QAPMIDLC QAPMSTNL
QAPMBSC QAPMIOPD QAPMSTNY
QAPMBUS QAPMJOBS QAPMSYS
QAPMCIOP QAPMLAPD QAPMTJOB
QAPMCONF QAPMLIOP QAPMTSK
QAPMDBMON QAPMMIOP QAPMX25
QAPMDDI QAPMPOOL QAPTLCKD
QAPMDIOP QAPMRESP QTRDMPT
QAPMDISK QAPMRWS QTRIDX
QAPMDMPT QAPMSAP QTRJOBT
QAPMECL QAPMSBSD QTRJSUM
QAPMETH QAPMSNA QTRTSUM
QAPMFRLY QAPMSNADS

Copy Performance Data

Use option 2 (Copy performance data) on the Manage Performance Data display to
make copies of performance data members. When you choose option 2, the Select
Performance Member display appears.

15-2 Performance Tools V4R2

à ð
Select Performance Member

Library . . . . . . QPFRDATA__

Type option, press Enter.

1=Select

Option Member Text Date Time

_ FRIDAY Performance Data for Friday 1ð/27/95 1ð:ð5:46
_ THURSDAY Performance Data for Thursday 1ð/26/95 12:ðð:34
1 WEDNESDAY Performance Data for Wednesday 1ð/25/95 13:5ð:15
1 TUESDAY Performance Data for Tuesday 1ð/24/95 13:55:ð8
1 MONDAY Performance Data for Monday 1ð/23/95 16:25:39
_ TESTRUN Test run of system 1ð/19/95 2ð:31:42
_ Q95291ð958 1ð/18/95 ð9:58:45
_ Q9529ð2ðð9 1ð/17/95 2ð:ð9:23

F3=Exit F12=Cancel F15=Sort by member F16=Sort by text

F19=Sort by date/time
(C) COPYRIGHT IBM CORP. 1981, 1995.
á ñ
The members that appear on this display are those used on the Start Performance
Monitor (STRPFRMON) command for the keyword MBR when data was collected.
To copy a member or members from the list, type a 1 (Select) next to the appro-
priate member(s) and press the Enter key.

The Copy Performance Data Member display appears.

à ð
Copy Performance Data Member

Type choices, press Enter.

--------Copy From-------- --------Copy To--------

Member Library Member Library
MONDAY QPFRDATA MONDAY NEWLIB
TUESDAY QPFRDATA TUESDAY NEWLIB
WEDNESDAY QPFRDATA WEDNESDAY NEWLIB
Bottom

F3=Exit F12=Cancel

á ñ
This display shows you the members you selected to copy and where they are to
be copied to. For each member listed, type the name of the new member and the
library that contains it in the Copy To entries of the display, and then press the

Chapter 15. Managing the Performance Data—Agent Feature 15-3

 Enter key. When the copy completes, you have exact copies of the old perform-
ance members in the new performance members for the following files:

QAPMAPPN QAPMFRLY QAPMSAP

QAPMASYN QAPMHDLC QAPMSBSD
QAPMBSC QAPMHDWR QAPMSNA
QAPMBUS QAPMIDLC QAPMSNADS
QAPMCIOP QAPMIOPD QAPMSTND
QAPMCONF QAPMJOBS QAPMSTNE
QAPMDBMON QAPMLAPD QAPMSTNL
QAPMDDI QAPMLIOP QAPMSTNY
QAPMDIOP QAPMMIOP QAPMSYS
QAPMDISK QAPMPOOL QAMPTJOB
QAPMDMPT QAPMRESP QAPMTSK
QAPMECL QAPMRWS QAPMX25
QAPMETH

Convert Performance Data (CVTPFRDTA) Command

Use option 4 (Convert performance data) on the Configure and Manage Tools
display.

When you select option 4, the Convert Performance Data (CVTPFRDTA) display
appears. You can also use the CVTPFRDTA command to select the CVTPFRDTA
display.

à ð
Convert Performance Data (CVTPFRDTA)

Type choices, press Enter.

From library . . . . . . . . . . ___________ Name

To library . . . . . . . . . . . ___________ Name
Job Description . . . . . . . . \USRPRF____ Name, \USRPRF, \NONE
Library . . . . . . . . . . . ___________ Name, \LIBL, \CURLIB

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

á ñ
The Convert Performance Data (CVTPFRDTA) command converts performance
data from the previous release to the formats needed to be processed by the
current release of the performance measurement/analysis tools. First, the release
level on which the data was collected is determined. Then, all members of all files
that need conversion are converted to the appropriate format.

15-4 Performance Tools V4R2

The following files must be present for the conversion to take place:

QAPMCIOP QAPMDISK QAPMPOOL

QAPMCONF QAPMJOBS QAPMRESP
QAPMDIOP QAPMLIOP QAPMSYS

The following files are copied, or converted if necessary, if they are present:

QACPCNFG QAPMASYN QAPMIDLC

QACPGPHF QAPMBSC QAPMLAPD
QACPPROF QAPMBUS QAPMMIOP
QACPRESP QAPMDMPT QAPMSBSD
QAITMON QAPMECL QAPMTSK
QAPGHSTD QAPMETH QAPMX25
QAPGHSTI QAPMHDLC QAPTAPGP
QAPGPKGF

The conversion can be done in the library in which the current data resides, or in a
different library. If the conversion is done in the same library, the current data is
replaced by the new data. If the conversion is done in a different library, the new
data exists in the new library while the current data continues to exist in the current
library.
Note: If a different library is specified for the new data, those files in the current
library that do not need conversion are copied to the new library.

To convert performance data collected prior to the current release, complete the
following items on the display.
From library Specifies the library that contains the data being converted.
To library Specifies the library that contains the converted data.
Job Description Specifies the job description used to submit the file-conversion job
for batch processing.

The possible job description values are:

*USRPRF The job description defined for the submitting job’s user profile.
job-description-name Specify the name of the job description to be used.
*NONE A batch job is not submitted. Processing continues interactively while the
user waits.
Note: The user’s work station is not available for other use during this time, which
can be significant for long jobs.

The possible library values are:

*LIBL The library list is used to locate the job description.
*CURLIB The current library for the job is used to locate the job description. If no
current library entry exists in the library list, QGPL is used.
library-name The library where the job description is located.

Chapter 15. Managing the Performance Data—Agent Feature 15-5

| Convert Performance Thread Data (CVTPFRTHD) Command
| The Convert Performance Thread Data (CVTPFRTHD) command converts perform-
| ance data records collected by the STRPFRMON command. The specified member
| of database file QAPMJOBS contains records with thread-level performance data.
| You can use the CVTPFRTHD command to convert the data and write the records
| to a member in file QAPMTJOB. The output file member contains records with job-
| level performance data which are a total of the performance information for all
| threads running within the job.

| à ð
| Convert Pfr Thread Data (CVTPFRTHD)

| Type choices, press Enter.

| Member . . . . . . . . . . . . . Name
| Library . . . . . . . . . . . . QPFRDATA Name
| Replace . . . . . . . . . . . . \YES \YES, \NO

| Bottom
| F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
| F24=More keys

| á ñ

15-6 Performance Tools V4R2

Appendix A. Performance Tools CL Commands
The following commands are part of the Performance Tools Manager feature. You
must have the Manager feature of the Performance Tools licensed program
installed to use these commands.

ANZACCGRP DLTGPHFMT PRTPOLRPT

ANZDBF DLTGPHPKG PRTRSCRPT
ANZDBFKEY DLTHSTDTA PRTSYSRPT
ANZPFRDTA DLTPFRDTA PRTTNSRPT
ANZPGM DSPACCGRP PRTTRCRPT
CHGFCNARA DSPHSTGPH STRJOBTRC
CHGGPHFMT DSPPFRDTA STRPFRG
CHGGPHPKG DSPPFRGPH STRPFRT
CPYFCNARA ENDJOBTRC WRKFCNARA
CPYGPHFMT PRTACTRPT WRKSYSACT
CPYGPHPKG PRTCPTRPT
CPYPFRDTA PRTJOBRPT
CRTFCNARA PRTJOBTRC
CRTGPHFMT PRTLCKRPT
CRTGPHPKG
CRTHSTDTA
DLTFCNARA

The following commands are part of the Performance Tools Agent feature as well
as the Performance Tools Manager feature.

ANZPFRDTA DLTHSTDTA PRTPEXRPT

CPYPFRDTA DLTPFRDTA STRPFRT
CRTHSTDTA DSPPFRDTA WRKSYSACT

The following commands relate to the performance explorer tool and are part of the
OS/400 product (with the exception of the PRTPEXRPT command, which is part of
the Performance Tools licensed program). You do not have to have the Perform-
ance Tools licensed program installed to use these commands.

ADDPEXDFN ENDPEX
CHGPEXDFN RMVPEXDFN
DLTPEXDTA STRPEX

The following commands are part of the OS/400 product and are described in the
CL Reference and the Programming Reference Summary book. You do not have to
have the Performance Tools licensed program installed to use these commands.

ADDPFRCOL DMPTRC STRPFRCOL

CHGPFRCOL ENDPFRCOL STRPFRMON
CVTPFRDTA ENDPFRMON WRKPFRCOL
CVTPFRTHD

The following commands are BEST/1 Capacity Planning commands. They are part
of the Performance Tools licensed program and are described in the BEST/1
Capacity Planning Tool book. You must have the Performance Tools licensed
program installed to use these commands.
Ÿ ANZBESTMDL (Manager feature only)

 Copyright IBM Corp. 1998 A-1

 ADDPEXDFN

Ÿ CRTBESTMDL (Manager and Agent feature)

Ÿ DLTBESTMDL (Manager and Agent feature)
Ÿ STRBEST (Manager feature only)

See Table D-2 on page D-2 for a table showing all performance-related commands
and whether they are part of OS/400, part of the Manager feature of Performance
Tools, or part of the Agent feature of Performance Tools.

ADDPEXDFN (Add Performance Explorer Definition) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──ADDPEXDFN──DFN(──definition-name──)─── ─┬────────────────────────┬─────5
│ ┌─\STATS───┐ │
└─TYPE(──┼─\TRACE───┼──)─┘
└─\PROFILE─┘
5──┬──────────────────────────┬───────────────────────────────────────────5
│ ┌─\───────────┐ │
└─JOB(──┼─\ALL────────┼──)─┘
│ ┌──
─────────┐ │
└──6─┤ Job ├─┴─┘
5──┬────────────────────────────────────────────┬─────────────────────────5
| │ ┌─\NONE────────────────────────┐ │
| └─TASK(──┼─\ALL─────────────────────────┼──)─┘
| │ ┌──
────────────────────────┐ │
| └──6─┬─task-name──────────┬─┴──
(2)
─┘
| └─generic\-task-name─┘
5──┬──────────────────────────────────────────────────────────────┬───────5
│ ┌──
─────────────────────────────────────────────┐ │
│ │ ┌─\LIBL/───┐ │ │
(3) 6 (4)
└─PGM(─── ───┼──────────┼──program-name──┤ PGM Details ├─┴── ──)─┘
└─library/─┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\FLAT─┐ │
(5)
└─DTAORG(─── ─┴─\HIER─┴──)─┘
5──┬────────────────────────────────────────────┬─────────────────────────5
│ ┌─1ðððð───────────────────┐ │
(6, 7)
└─MAXSTG(──── ─┴─maximum-K-bytes-storage─┴──)─┘
5──┬─────────────────────────────┬──┬───────────────────────────┬─────────5
│ ┌─\STOPTRC─┐ │ │ ┌─\YES─┐ │
(8) (9)
└─TRCFULL(─── ─┴─\WRAP────┴──)─┘ └─INCDEPJOB(─── ─┴─\NO──┴──)─┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\YES─┐ │
(10)
└─MRGJOB(──── ─┴─\NO──┴──)─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
(11, 12)
└─INTERVAL(────── ───milliseconds────)─┘

A-2 Performance Tools V4R2

 ADDPEXDFN

5──┬────────────────────────────────────┬──┬─────────────────────────┬────5
| │ ┌──
───────────────┐ │ │ ┌─\NO──┐ │
(13)
| │ │ ┌─\CALLRTN──┐ │ │ └─SLTEVT(──── ─┴─\YES─┴──)─┘
| └─TRCTYPE(──┬──6─┼─\BASIC────┼─┴─┬──)─┘
| │ ├─\DSKIO1───┤ │
| │ ├─\DSKIO2───┤ │
| │ ├─\DSKSVR───┤ │
| │ ├─\DSKSTG───┤ │
| │ ├─\VRTADR───┤ │
| │ ├─\PGMACT───┤ │
| │ ├─\FILEOPEN─┤ │
| │ ├─\PRFDTA───┤ │
| │ └─\TASKSWT──┘ │
| └─\SLTEVT───────────┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─\BLANK────────┐ │
└─TEXT(──┴─'description'─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
| │ ┌─\ALL────────────────────────────┐ │
(14)
| └─MCHINST(──── ─┼─\NONE───────────────────────────┼──)─┘
| │ ┌──
──────────────────────────┐ │
| └──6─machine-instruction-name─┴───
(15)
─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(16)
| └─BASEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(17)
| └─PGMEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(18)
| └─STGEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(19)
| └─DSKEVT(──── ─┼─\ALL─────────────────┼──)─┘
| ├─\ALLSTR──────────────┤
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
| │ ┌─\NONE────────────────┐ │
(20)
| └─FAULTEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(21)
| └─JOBEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(22)
| └─LCKEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘

Appendix A. Performance Tools CL Commands A-3

 ADDPEXDFN

5──┬─────────────────────────────────────────┬────────────────────────────5
| │ ┌─\NONE────────────────┐ │
(23)
| └─SAREVT(──── ─┼─\ALL─────────────────┼──)─┘
| ├─\ALLSTR──────────────┤
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬────────────────────────────────────────────┬─────────────────────────5
| │ ┌─\NONE────────────────┐ │
(24)
| └─DSKSVREVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
5──┬─────────────────────────────────────────┬───────────────────────────5%
| │ ┌─\NONE────────────────┐ │
(25)
| └─JVAEVT(──── ─┼─\ALL─────────────────┼──)─┘
| │ ┌──
──────────────────┐ │
| └──6─event-identifier─┴─┘
Job:
┌─\ALL/───────┐ ┌─\ALL/───────────────┐
(1)
├──┼─────────────┼──┼─────────────────────┼──┬─job-name──────────┬─── ─────┤
└─job-number/─┘ ├─generic\-user-name/─┤ └─generic\-job-name─┘
└─user-name/──────────┘
PGM Details:
┌─\ALL────────┐ ┌─\ALL──────┐ ┌─\PGM────┐ ┌─4─────────┐
├──────┴─module-name─┴──┴─procedure─┴──┴─\SRVPGM─┴──┴─pane-size─┴─────────┤
Notes:
P All parameters preceding this point can be specified in positional form.
1 A maximum of 10 repetitions.

| 2 A maximum of 10 repetitions.

3 Only valid when TYPE(*PROFILE) is specified.

4 A maximum of 16 repetitions.

5 Only valid when TYPE(*STATS) is specified.

6 Only valid when TYPE(*TRACE) is specified.

7 Valid values range from 1K to 4,000,000 K (4 gigabytes).

8 Only valid when TYPE(*TRACE) is specified.

9 This parameter is not valid if JOB(*ALL) is specified.

10 Only valid when DTAORG(*FLAT) and TYPE(*STATS) are specified.

11 Only valid when TYPE(*PROFILE) or TYPE(*TRACE) is specified.

12 Valid values range from 1 to 100 milliseconds.

| 13 Only valid when TYPE(*TRACE) is specified.

| 14 Only valid when SLTEVT(*YES) is specified.

| 15 A maximum of 50 repetitions.

| 16 Only valid when SLTEVT(*YES) is specified.

| 17 Only valid when SLTEVT(*YES) is specified.

| 18 Only valid when SLTEVT(*YES) is specified.

| 19 Only valid when SLTEVT(*YES) is specified.

| 20 Only valid when SLTEVT(*YES) is specified.

| 21 Only valid when SLTEVT(*YES) is specified.

| 22 Only valid when SLTEVT(*YES) is specified.

| 23 Only valid when SLTEVT(*YES) is specified.

| 24 Only valid when SLTEVT(*YES) is specified.

| 25 Only valid when SLTEVT(*YES) is specified.

A-4 Performance Tools V4R2

 ANZACCGRP

Purpose
The Add Performance Explorer Definition (ADDPEXDFN) command adds a new
performance explorer definition to the system. Each definition is stored as a
member in the QAPEXDFN file in library QUSRSYS. A performance explorer defi-
nition identifies the performance data that is to be collected during a performance
explorer session. A session can be started using the STRPEX (Start Performance
Explorer) command. When starting a new session, a performance explorer defi-
nition name must be provided.

Examples
Example 1: Using TYPE(*TRACE)
ADDPEXDFN DFN(TEST1) TYPE(\TRACE)
JOB(\) MAXSTG(5ððð)

This command adds a new performance definition named TEST1, which will result
in a member named TEST1 being added to file QAPEXDFN in library QUSRSYS.
When this definition is used to start a performance explorer session (STRPEX
command), detailed trace information will be collected for the job that invoked the
STRPEX command. A maximum of 5000 kilobytes of trace data will be collected.
When the trace record storage area is full no more trace records will be collected.

Example 2: Using TYPE(*PROFILE)

ADDPEXDFN DFN(TEST2) TYPE(\PROFILE)
PGM((MYLIB/MYSRVPGM1 \ALL \ALL \SRVPGM))

This command adds a new performance explorer definition named TEST2. When
this definition is used to start a performance explorer session (STRPEX command),
performance profile information for service program MYSRVPGM1 in library MYLIB
will be collected.

ANZACCGRP (Analyze Access Group) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──ANZACCGRP──┬──────────────────────────┬───────────────────────────────5
│ ┌─QAPAGDTA────┐ │
└─MBR(──┴─member-name─┴──)─┘
5──┬───────────────────────────┬──┬───────────────────────────────┬───────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\BLANK─────────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─TITLE(──┴─'report-title'─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─ANZACCGRP─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘

Appendix A. Performance Tools CL Commands A-5

 ANZDBF

Purpose
The Analyze Access Group (ANZACCGRP) command produces a report that sum-
marizes the Process Access Group (PAG) data collected earlier with the Display
Access Group (DSPACCGRP) command. The report is useful in examining the
process access group data from a large number of jobs.

| The environment section of the report shows a summary of all job types, the
| average number of files, duplicate files, display files, I/O counts, and process-
| access group size for the different job types. The job section shows information for
| each selected job. The file section shows, for each open file, the file name and
| type, the number of jobs using the file, the number of opens, the I/O count, and the
| average ODP (Open Data Path) size.

The report output is written to the printer file QSYSPRT.

Example
ANZACCGRP

This command produces a report from the PAG data previously stored in the
default location, member QAPAGDTA of QPFRDATA/QAPTPAGD, by the Display
Access Group (DSPACCGRP) command.

ANZDBF (Analyze Database File) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌──
──────────────┐
55──ANZDBF──LIBL(───6─library-name─┴──
(1) (P)
──)─── ─┬───────────────────────┬─────5
│ ┌─ANZDBF───┐ │
└─JOB(──┴─job-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
1 A maximum of 10 repetitions
P All parameters preceding this point can be specified in positional form.

Purpose
The Analyze Database File (ANZDBF) command produces two reports that show
the physical and logical files in a set of libraries and the relationships between the
files. It saves the information in a database file for further analysis by the Analyze
Database File Keys (ANZDBFKEY) command. Both reports (physical to logical file
relationships and logical to physical file relationships) are written to the printer file
QPPTANZD (two printer files with the same name are produced; one printer file
contains summary data; the other contains detail data). The data is saved in
member QAPTAZDR of the database file QPFRDATA/QAPTAZDR.

A-6 Performance Tools V4R2

 ANZPFRDTA

Example
ANZDBF LIBL(APDTA ARDTA)

This command produces reports showing the relationships for all files in the
Accounts Payable (APDTA) and Accounts Receivable (ARDTA) data libraries.

ANZDBFKEY (Analyze Database File Keys) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──ANZDBFKEY──┬─────────────────────────┬────────────────────────────────5
│ ┌─\NUMLF────┐ │
└─FILE(──┴─file-name─┴──)─┘
5──┬───────────────────────────┬──┬────────────────────────┬──────────────5
│ ┌─5──────────┐ │ │ ┌─ANZDBFKEY─┐ │
└─NUMLF(──┴─file-count─┴──)─┘ └─JOB(──┴─job-name──┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘

Purpose
The Analyze Database File Keys (ANZDBFKEY) command produces, from the data
created by the ANZDBF command, two reports showing the key structure of the
database files.

One report is written to the printer file QPPTALNK, the other to QPPTKEYMTX.

Example
ANZDBFKEY FILE(\NUMLF) NUMLF(2)

This command produces reports on the keys for all files that reference physical files
with at least two associated logical files.

ANZPFRDTA (Analyze Performance Data) Command

Format
Job: I Pgm: I REXX: I Exec

Appendix A. Performance Tools CL Commands A-7

ANZPGM

55──ANZPFRDTA──┬──────────────────────────┬───────────────────────────────5
│ ┌─\SELECT─────┐ │
└─MBR(──┴─member-name─┴──)─┘
5──┬───────────────────────────┬──┬─────────────────────┬─────────────────5
│ ┌─QPFRDATA─────┐ │ └─PERIOD(──┤ Period ├─┘
└─LIB(──┴─library-name─┴──)─┘
5──┬────────────────────────┬──┬──────────────────────────┬──────────────5%
│ ┌─\──────┐ │ │ ┌─\ALL────┐ │
└─OUTPUT(──┴─\PRINT─┴──)─┘ └─DTATYPE(──┴─\SAMPLE─┴──)─┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┼─\SELECT────┼──┼────────────┼──)──┬──────────────────────────────────┬──┤
└─start-time─┘ └─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘

Purpose
The Analyze Performance Data (ANZPFRDTA) command produces recommenda-
tions to improve the performance of the user's system. In the interactive mode, you
can request that the system make the recommended changes. In the batch mode,
the recommended changes are printed, and you must then must enter the indi-
vidual commands to make the recommended changes.

Example
ANZPFRDTA

This command provides recommendations for improving the performance of the

system.

ANZPGM (Analyze Program) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌──
──────────────┐
55──ANZPGM──LIBL(───6─library-name─┴──
(1) (P)
──)─── ─┬───────────────────────┬─────5
│ ┌─ANZPGM───┐ │
└─JOB(──┴─job-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
1 A maximum of 10 repetitions
P All parameters preceding this point can be specified in positional form.

A-8 Performance Tools V4R2

 CHGFCNARA

Purpose
The Analyze Program (ANZPGM) command produces a report that shows the pro-
grams and files in a set of libraries and the relationships between them. Both
reports are written to the printer file QPPTANZP (two print files are produced with
the same name; one printer file contains summary data; the other contains detail
data).

Example
ANZPGM LIBL(APPGM ARPGM)

This command produces reports showing the program and file relationships for all
programs in the Accounts Payable (APPGM) and Accounts Receivable (ARPGM)
program libraries.

CHGFCNARA (Change Functional Area) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──CHGFCNARA──FCNARA(──functional-area-name──)───────────────────────────5
(P)
5──┬───────────────────────────┬── ──┬─────────────────────────────┬───────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\SAME─────────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─TEXT(──┼─\BLANK────────┼──)─┘
└─'description'─┘
5──┬──────────────────────────────────────────────────┬──────────────────5%
│ ┌─\SAME───────────────────────────────┐ │
└─JOB(──┼─\NONE───────────────────────────────┼──)─┘
│ ┌──
───────────────────────────────┐ │
└──6───┬─────────────┬──job name───┴──
(1)
─┘
└─ user-name/─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 A maximum of 250 repetitions

Purpose
The Change Functional Area (CHGFCNARA) command allows the user to change
functional areas on the system. Functional areas are used by Performance Tools
for reports and graphics. A functional area is a pre-defined list of job names and/or
user names that are to be included in a report or graph.

Examples
Example 1: Changing Functional Area to Three Entries
CHGFCNARA FCNARA(PERSONNEL)
JOB(DAN/\N MARCY/\N RANDY/QPG\)

This command changes the functional area PERSONNEL to three entries:

Ÿ The user DAN.
Ÿ The user MARCY.
Ÿ Any job beginning with QPG submitted by RANDY.

Appendix A. Performance Tools CL Commands A-9

CHGGPHFMT

The functional area is changed in the QPFRDATA library.

Example 2: Changing Functional Area to Four Entries

CHGFCNARA FCNARA('Performance Tools') LIB(RPFT)
JOB(PRT\ TERESA/\N KAREN/\N JIM/QPFRMON)

This command changes the functional area 'Performance Tools' to four entries:
Ÿ Any job beginning with PRT.
Ÿ The user TERESA.
Ÿ The user KAREN.
Ÿ Any QPFRMON job submitted by JIM.
The functional area is created in library RPFT.

CHGGPHFMT (Change Graph Format) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
55──CHGGPHFMT──GPHFMT(──┼───────────────┼──format-name──)─────────────────5
└─library-name/─┘
(P, K)
5──┬─────────────────────────────┬──── ────────────────────────────────────5
│ ┌─\SAME─────────┐ │
└─TEXT(──┼─\BLANK────────┼──)─┘
└─'description'─┘
5──┬──────────────────────────────┬───────────────────────────────────────5
│ ┌─\SAME─────────┐ │
└─TITLE(──┼─\BLANK────────┼──)─┘
├─\MBRTEXT──────┤
└─'graph-title'─┘
5──┬────────────────────────────────────┬─────────────────────────────────5
│ ┌─\SAME────────────┐ │
└─SUBTITLE(──┼─\BLANK───────────┼──)─┘
├─\MBRTEXT─────────┤
└─'graph-subtitle'─┘
5──┬───────────────────────────┬──┬─────────────────────────────┬─────────5
│ ┌─\SAME────┐ │ │ ┌─\SAME─────┐ │
└─GPHTYPE(──┼─\SURFACE─┼──)─┘ └─DATATYPE(──┼─\ALL──────┼──)─┘
├─\LINE────┤ ├─\FCNARA───┤
├─\CBAR────┤ ├─\JOBTYPE──┤
├─\FBAR────┤ ├─\PRIORITY─┤
└─\SCATTER─┘ ├─\IOP──────┤
├─\DISK─────┤
└─\CMNLINE──┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\SAME─┐ │
└─AREAFILL(──┼─\NO───┼──)─┘
└─\YES──┘
5──┬────────────────────────────────────────┬─────────────────────────────5
│ ┌─\SAME─────────────────┐ │
└─REFLINE(──┼─\NONE─────────────────┼──)─┘
└─reference-line-number─┘

A-10 Performance Tools V4R2

 CHGGPHFMT

5──┬──────────────────────┬──┬──────────────────────┬─────────────────────5
└─XAXIS(──┤ Xaxis ├──)─┘ └─YAXIS(──┤ Yaxis ├──)─┘
5──┬──────────────────────────┬───────────────────────────────────────────5
(1)
└─FCNARA(─── ─┤ Fcnara ├──)─┘
5──┬────────────────────────────────────────────────────────────────────┬──5
│ ┌─\SAME───────────────────────────────────────────┐ │
│ │ ┌──
───────────────────────────────────────────┐ │ │
(3)
└─JOBTYPE(─── ─┴──6─(──┬─\ALL───────┬──┬─────────────────┬──)─┴──
(4)
─┴──)─┘
├─\ALLINTER──┤ └─┤ More Values ├─┘
├─\ALLBATCH──┤
├─\ALLSYSTEM─┤
├─\ASJ───────┤
├─\BCH───────┤
├─\CA4───────┤
├─\DDM───────┤
├─\EVK───────┤
├─\INT───────┤
├─\MRT───────┤
├─\PCS───────┤
├─\PDJ───────┤
├─\PJ────────┤
├─\PTH───────┤
├─\RDR───────┤
├─\S36───────┤
├─\SBS───────┤
├─\SYS───────┤
├─\WTR───────┤
└─\OTHER─────┘
5──┬────────────────────────────────────────────────────────────┬────────5%
(5)
├─PRIORITY(─── ─┤ Priority ├──)───────────────────────────────┤
│ ┌─\SAME─────────────────────────────────────┐ │
│ │ ┌───────────────────────────────────────┐ │ │
(7)
├─IOP(─── ─┴──6─(──┬─\AVG─┬──┬─────────────────┬──)─┴── (8)
─┴──)───┤
│ └─\MAX─┘ └─┤ More Values ├─┘ │
│ ┌─\SAME──────────────────────────────────────┐ │
│ │ ┌── ─────────────────────────────────────┐ │ │
(9)
├─DISK(─── ─┴──6─(──┬─\AVG─┬──┬─────────────────┬──)─┴───(10)
─┴──)─┤
│ └─\MAX─┘ └─┤ More Values ├─┘ │
(11)
├─CMNLINE(──── ─┤ Cmnline ├──)────────────────────────────────┤
│ ┌─\SAME───────────────┐ │
(13)
└─ALLDATA(──── ─┴─┬─────────────────┬─┴──)────────────────────┘
└─┤ More Values ├─┘
More Values:
├──┬────────────────────────────────────────────────────────────┬─────────┤
│ ┌─\DFT──────────┐ │
└─┼─\BLANK────────┼──┬─────────────────────────────────────┬─┘
(14)
└─'description'─┘ │ ┌─7──── ────────────┐ ┌─\─────────┐ │
└─┴─line-type-number─┴──┼───────────┼─┘
└─character─┘
Xaxis:
┌─\SAME─────┐
├──┼─\TIME─────┼──┬──────────────────────────────────────────────────┬────┤
├─\CPU──────┤ │ ┌─\SAME──────────┐ ┌─\SAME────────────────────┐ │
├─\TNS──────┤ └─┼─\DFT───────────┼──┼──────────────────────────┼─┘
├─\NBRTNS───┤ ├─\BLANK─────────┤ ├─\AUTO────────────────────┤
├─\RSP──────┤ └─'x-axis-title'─┘ └─start-number──end-number─┘
├─\SYNCIO───┤
├─\NBRSYNC──┤
├─\ASYNCIO──┤
├─\NBRASYNC─┤
├─\TOTDSKIO─┤
└─\NBRDSKIO─┘

Appendix A. Performance Tools CL Commands A-11

CHGGPHFMT

Yaxis:
┌─\SAME──────┐
├──┼─\CPU───────┼──┬──────────────────────────────────────────────────┬───┤
├─\TNS───────┤ │ ┌─\SAME──────────┐ ┌─\SAME────────────────────┐ │
├─\NBRTNS────┤ └─┼─\DFT───────────┼──┼──────────────────────────┼─┘
├─\RSP───────┤ ├─\BLANK─────────┤ ├─\AUTO────────────────────┤
├─\SYNCIO────┤ └─'y-axis-title'─┘ └─start-number──end-number─┘
├─\NBRSYNC───┤
├─\ASYNCIO───┤
├─\NBRASYNC──┤
├─\TOTDSKIO──┤
├─\NBRDSKIO──┤
├─\CMNIOP────┤
├─\DSKIOP────┤
├─\LWSIOP────┤
├─\MFCIOP────┤
├─\MFDIOP────┤
├─\DSKARM────┤
├─\PCTDSKOCC─┤
└─\CMNLINE───┘
Fcnara:
┌─\SAME─────────────────────────────────────────────────────┐
│ ┌──
─────────────────────────────────────────────────────┐ │
├──┴──6─(──┬─\OTHER───────────────┬──┬─────────────────┬──)─┴──
(2)
─┴──────────┤
└─functional-area-name─┘ └─┤ More Values ├─┘
Priority:
┌─\SAME─────────────────────────────────────────────────────────┐
│ ┌──
─────────────────────────────────────────────────────────┐ │
├──┴──6─(──┬─\ALL─────────────────────┬──┬─────────────────┬──)─┴──
(6)
─┴──────┤
├─\OTHER───────────────────┤ └─┤ More Values ├─┘
└─lower-limit──upper-limit─┘
Cmnline:
┌─\SAME──────────────────────────────────────────────────────────┐
│ ┌──
─────────────────────────────────────────────────────────┐ │
├──┴──6─(──┬─\MAX─────────────────────┬──┬─────────────────┬──)─┴───
(12)
─┴─────┤
└─communications-line-name─┘ └─┤ More Values ├─┘
Notes:
P All parameters preceding this point can be specified in positional form.

K All parameters preceding this point are key parameters.

1 Valid only when DATATYPE(*FCNARA) is specified

2 A maximum of 16 repetitions

3 Valid only when DATATYPE(*JOBTYPE) is specified

4 A maximum of 16 repetitions

5 Valid only when DATATYPE(*PRIORITY) is specified

6 A maximum of 16 repetitions

7 Valid only when *DATATYPE(*IOP) is specified

8 A maximum of 2 repetitions

9 Valid only when DATATYPE(*DISK) is specified.

10 A maximum of 2 repetitions

11 Valid only when DATATYPE(*CMNLINE) is specified.

12 A maximum of 16 repetitions

13 Valid only when DATATYPE(*ALL) is specified.

14 Solid line

A-12 Performance Tools V4R2

 CHGGPHPKG

Purpose
The Change Graph Format (CHGGPHFMT) command changes a graph format
used to display performance and historical graphs that are created from perform-
ance data members.

Example
CHGGPHFMT GRAPH(FORMAT1) DATATYPE(\ALL)
ALLDATA(ABCCOMPANY \DFT 7)

This command changes the graph format named FORMAT1 in the QPFRDATA
library and groups all the data together. The data legend description is
ABCCOMPANY, and it is represented on the graph with a solid line.

CHGGPHPKG (Change Graph Package) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P, K)
55──CHGGPHPKG──GPHPKG(──┼───────────────┼──package-name──)───── ───────────5
├─\CURLIB/──────┤
└─library-name/─┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─\SAME─────────┐ │
└─TEXT(──┼─\BLANK────────┼──)─┘
└─'description'─┘
5──┬───────────────────────────────────┬─────────────────────────────────5%
│ ┌─\SAME─────────────┐ │
(1)
└─GPHFMT(──┼─\SELECT─── ────────┼──)─┘
│ ┌──
─────────────┐ │
└──6─format-name─┴──(2)
─┘
Notes:
P All parameters preceding this point can be specified in positional form.
K All parameters preceding this point are key parameters.

1 *SELECT is not valid in batch.

2 A maximum of 25 repetitions

Purpose
The Change Graph Package (CHGGPHPKG) command changes a graph package
to include one or more graph formats.

Examples
Example 1: Changing Graph Package to Contain Three Formats
CHGGPHPKG GPHPKG(EXAMPLE) TEXT('THIS IS AN
EXAMPLE') GPHFMT(GPH1 GPH9 GPH12)

This command changes a graph package called EXAMPLE to contain three

formats, GPH1, GPH9, and GPH12. This package is located in the default library,
QPFRDATA.

Appendix A. Performance Tools CL Commands A-13

CHGPEXDFN

Example 2: Changing Graph Package to Contain Two Formats

CHGGPHPKG GPHPKG(MYLIB/MYPKG)
TEXT('MY PACKAGE') GPHFMT(MYGPH1 MYGPH2)

This command changes a graph package called MYPKG to contain the formats of
MYGPH1 and MYGPH2. MYPKG is located in the MYLIB library.

CHGJOBTYP (Change Job Type) Command

Format
Job: I Pgm: I REXX: I Exec

55──CHGJOBTYP──MBR(──performance-data-member-name──)──────────────────────5
(P)
5──┬────────────────────────────────────────────┬── ──────────────────────5%
│ ┌─QPFRDATA──────────────────────┐ │
└─LIB(──┴─performance-data-library-name─┴──)─┘
Note:
P All parameters preceding this point can be specified in positional form.

Purpose
The Change Job Type (CHGJOBTYP) command allows you to change the job type
for jobs that appear on the reports you produce using the Print Transaction Report
(PRTTNSRPT) command. With this command you can change the characteristics
of a job (for example, from batch to interactive), correct a missing job type, or
assign a job type.

Examples
Example 1: Start a New Session
CHGJOBTYP MBR(TEST)
LIB(QPFRDATA)

This command changes MBR TEST from a batch job to an interactive job.

CHGPEXDFN (Change Performance Explorer Definition) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P, K)
55──CHGPEXDFN──DFN(──definition-name──)───── ─┬────────────────────────┬───5
│ ┌─\SAME────┐ │
└─TYPE(──┼─\STATS───┼──)─┘
├─\TRACE───┤
└─\PROFILE─┘
5──┬────────────────────────────┬─────────────────────────────────────────5
│ ┌─\SAME─────────┐ │
└─JOB(──┼─\─────────────┼──)─┘
├─\ALL──────────┤
│ ┌──
─────────┐ │
└──6─┤ Job ├─┴──
(2)
─┘

A-14 Performance Tools V4R2

 CHGPEXDFN

5──┬────────────────────────────────────────────┬─────────────────────────5
| │ ┌─\SAME────────────────────────┐ │
| └─TASK(──┼─\NONE────────────────────────┼──)─┘
| ├─\ALL─────────────────────────┤
| │ ┌──
────────────────────────┐ │
| └──6─┬─task-name──────────┬─┴──
(3)
─┘
| └─generic\-task-name─┘
5──┬────────────────────────────────────────────────────────────────┬─────5
│ ┌─\SAME───────────────────────────────────────────┐ │
│ │ ┌──
─────────────────────────────────────┐ │ │
│ │ │ ┌─\LIBL/───┐ │ │ │
(4)
└─PGM(─── ─┴─(───6─┼──────────┼──program-name──┤ Pgm ├─┴──
(5)
──)─┴──)─┘
└─library/─┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\SAME─┐ │
(6)
└─DTAORG(─── ─┼─\FLAT─┼──)─┘
└─\HIER─┘
5──┬────────────────────────────────────────────┬─────────────────────────5
│ ┌─\SAME───────────────────┐ │
(7, 8)
└─MAXSTG(──── ─┴─maximum-K-bytes-storage─┴──)─┘
5──┬─────────────────────────────┬──┬─────────────────────────────┬───────5
│ ┌─\SAME────┐ │ │ ┌─\SAME─┐ │
(9) (10)
└─TRCFULL(─── ─┼─\STOPTRC─┼──)─┘ └─INCDEPJOB(──── ─┼─\YES──┼──)─┘
└─\WRAP────┘ └─\NO───┘
5──┬──────────────────────────┬───────────────────────────────────────────5
│ ┌─\SAME─┐ │
(11)
└─MRGJOB(──── ─┼─\YES──┼──)─┘
└─\NO───┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\SAME─────────────┐ │
(12, 13)
└─INTERVAL(────── ─┴─sampling-interval─┴──)─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
| │ ┌─\SAME───────────┐ │
(14) (15)
| └─TRCTYPE(──── ─┼─\SLTEVT──── ─────┼──)─┘
| │ ┌──
─────────────┐ │
| └──6┬─\CALLRTN──┬┴─┘
| ├─\BASIC────┤
| ├─\DSKIO1───┤
| ├─\DSKIO2───┤
| ├─\DSKSVR───┤
| ├─\DSKSTG───┤
| ├─\VRTADR───┤
| ├─\PGMACT───┤
| ├─\FILEOPEN─┤
| ├─\PRFDTA───┤
| └─\TASKSWT──┘
5──┬─────────────────────────────┬───────────────────────────────────────5%
│ ┌─\SAME─────────┐ │
└─TEXT(──┼─\BLANK────────┼──)─┘
└─'description'─┘
Job:
┌─\ALL/───────┐ ┌─\ALL/───────────────┐
├──┼─────────────┼──┼─────────────────────┼──┬─job-name──────────┬────────┤
└─job-number/─┘ ├─generic\-user-name/─┤ └─generic\-job-name─┘
└─user-name/──────────┘
Pgm:
┌─\ALL────────┐ ┌─\ALL──────┐ ┌─\PGM────┐ ┌─4─────────┐
├──┼─\PGM────────┼──┴─procedure─┴──┴─\SRVPGM─┴──┴─pane-size─┴─────────────┤
└─module-name─┘

Appendix A. Performance Tools CL Commands A-15

 CHGPEXDFN

Notes:
1 Only valid when TYPE(*STATS) or TYPE(*TRACE) is specified.

P All parameters preceding this point can be specified in positional form.

K All parameters preceding this point are key parameters.

2 A maximum of 10 repetitions.

| 3 A maximum of 10 repetitions.

4 Only valid when TYPE(*PROFILE) is specified.

5 A maximum of 16 repetitions.

6 Only valid when TYPE(*STATS) is specified.

7 Valid values range from 1K to 4000000K (4 gigabytes).

8 Only valid when TYPE(*TRACE) is specified.

9 Only valid when TYPE(*TRACE) is specified.

10 This parameter is not valid if JOB(*ALL) is specified.

11 Only valid when DTAORG(*FLAT) and TYPE(*STATS) are specified.

12 Only valid when TYPE(*PROFILE) or TYPE(*TRACE) is specified.

13 Valid values range from 1 to 100 milliseconds.

| 14 Only valid when TYPE(*TRACE) is specified.

| 15 TRCTYPE(*SLTEVT) valid only when SLTEVT(*YES) specified on ADDPEXDFN

| command.

Purpose
The Change Performance Explorer Definition (CHGPEXDFN) command changes
an existing performance explorer definition. The current values stored in an existing
definition are inserted when the CHGPEXDFN command is prompted. Each defi-
nition is stored as a member in the QAPEXDFN file in library QUSRSYS. A per-
formance explorer definition identifies the performance data to be collected during a
performance explorer session. A session can be started using the Start Perform-
ance Explorer (STRPEX) command. When starting a new session, a performance
explorer definition name must be provided.

Examples
Example 1: Using TYPE(*TRACE)
CHGPEXDFN DFN(TEST1) TYPE(\TRACE)
JOB(\) MAXSTG(5ððð)

This command changes the performance definition named TEST1. When this defi-
nition is used to start a performance explorer session (STRPEX command),
detailed trace information will be collected for the job that invoked the STRPEX
command. A maximum of 5000 kilobytes of trace data will be collected. When the
trace record storage area is full no more trace records will be collected.

Example 2: Using TYPE(*PROFILE)

CHGPEXDFN DFN(TEST2) TYPE(\PROFILE)
PGM((MYLIB/MYSRVPGM1 \ALL \ALL \SRVPGM))

This command changes the performance explorer definition named TEST2. When
this definition is used to start a performance explorer session (STRPEX command),
performance profile information for service program MYSRVPGM1 in library MYLIB
will be collected.

A-16 Performance Tools V4R2

 CPYGPHFMT

CPYFCNARA (Copy Functional Area) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──CPYFCNARA──FROMFCNARA(──functional-area-name──)───────────────────────5
5──┬───────────────────────────────┬──────────────────────────────────────5
│ ┌─QPFRDATA─────┐ │
└─FROMLIB(──┴─library-name─┴──)─┘
5──┬────────────────────────────────────────┬─────────────────────────────5
│ ┌─\FROMFCNARA──────────┐ │
└─TOFCNARA(──┴─functional-area-name─┴──)─┘
(P)
5──┬─────────────────────────────┬── ──┬───────────────────────┬──────────5%
│ ┌─\FROMLIB─────┐ │ │ ┌─\NO──┐ │
└─TOLIB(──┴─library-name─┴──)─┘ └─REPLACE(──┴─\YES─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Copy Functional Area (CPYFCNARA) command allows the user to copy a
functional area over to a new functional area. Functional areas are used for Per-
formance Tools reports and graphics. A functional area is a pre-defined list of job
names and user names that are included in a report or graph.

Examples
Example 1: Copying in the Same Library
CPYFCNARA FROMFCNARA(PERSONNEL) TOFCNARA(MIKE)

This command copies the functional area PERSONNEL to the functional area
MIKE. Both functional areas are in the QPFRDATA library.

Example 2: Copying to a Different Library

CPYFCNARA FROMFCNARA('Performance Tools')
TOFCNARA(MIKE) TOLIB(USRLIB)

This command copies the functional area 'Performance Tools' to the functional area
MIKE in library USRLIB.

CPYGPHFMT (Copy Graph Format) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P)
55──CPYGPHFMT──FROMFMT(──┼───────────────┼──format-name──)─── ─────────────5
├─\CURLIB/──────┤
└─library-name/─┘

Appendix A. Performance Tools CL Commands A-17

CPYGPHPKG

5──┬───────────────────────────────────────────────┬──────────────────────5
│ ┌─\FROMLIB/─────┐ ┌─\FROMFMT────┐ │
└─TOFMT(──┼───────────────┼──┴─format-name─┴──)─┘
├─\CURLIB/──────┤
└─library-name/─┘
5──┬───────────────────────┬─────────────────────────────────────────────5%
│ ┌─\NO──┐ │
└─REPLACE(──┴─\YES─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Copy Graph Format (CPYGPHFMT) command copies an existing graph format
into a graph format specified by the user.

Examples
Example 1: Sending a Copy to the Default Library
CPYGPHFMT FROMFMT(MYFMT) TOFMT(YOURFMT)

This command makes a copy of MYFMT and sends it to YOURFMT in the

QPFRDATA library.

Example 2: Sending a Copy to a Specified Library

CPYGPHFMT FROMFMT(MYLIB/FMT1)
TOFMT(YOURLIB/FMT2)

This command makes a copy of FMT1 in the MYLIB library and sends it to FMT2 in
the YOURLIB library.

CPYGPHPKG (Copy Graph Package) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P)
55──CPYGPHPKG──FROMPKG(──┼───────────────┼──package-name──)─── ────────────5
├─\CURLIB/──────┤
└─library-name/─┘
5──┬────────────────────────────────────────────────┬─────────────────────5
│ ┌─\FROMLIB/─────┐ ┌─\FROMPKG─────┐ │
└─TOPKG(──┼───────────────┼──┴─package-name─┴──)─┘
├─\CURLIB/──────┤
└─library-name/─┘
5──┬───────────────────────┬─────────────────────────────────────────────5%
│ ┌─\NO──┐ │
└─REPLACE(──┴─\YES─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

A-18 Performance Tools V4R2

 CPYPFRDTA

Purpose
The Copy Graph Package (CPYGPHPKG) command copies an existing graph
package into a graph package specified by the user.

Examples
Example 1: Sending a Copy to the Default Library
CPYGPHPKG FROMPKG(MYPKG) TOPKG(YOURPKG)

This command makes a copy of MYPKG and sends it to YOURPKG in the

QPFRDATA library.

Example 2: Sending a Copy to a Specified Library

CPYGPHPKG FROMPKG(MYLIB/PKG1)
TOPKG(YOURLIB/PKG2)

This command makes a copy of PKG1 in the MYLIB library and sends it to PKG2
in the YOURLIB library.

CPYPFRDTA (Copy Performance Data) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──CPYPFRDTA──┬──────────────────────────────┬───────────────────────────5
(1)
│ ┌─\SELECT─── ──┐ │
└─FROMMBR(──┴─member-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬────────────────────────────┬────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\FROMMBR────┐ │
└─FROMLIB(──┴─library-name─┴──)─┘ └─TOMBR(──┴─member-name─┴──)─┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─\FROMLIB─────┐ │
└─TOLIB(──┴─library-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
1 *SELECT is not valid when submitting a batch job.
P All parameters preceding this point can be specified in positional form.

Purpose
The Copy Performance Data (CPYPFRDTA) command creates a copy of a per-
formance data member.

Appendix A. Performance Tools CL Commands A-19

CRTFCNARA

Examples
Example 1: Showing List of Performance Data Members
CPYPFRDTA

This command shows a display for selecting from all of the performance data
members in the QPFRDATA library. From this list, the user can select performance
data members to copy.

Example 2: Copying Data

CPYPFRDTA FROMMBR(MEMBER1) TOLIB(NEWLIB)

This command copies performance member MEMBER1 in library QPFRDATA to

library NEWLIB. The new copy in library NEWLIB keeps the name MEMBER1.

CRTFCNARA (Create Functional Area) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──CRTFCNARA──FCNARA(──functional-area-name──)───────────────────────────5
(P)
5──┬───────────────────────────┬── ──┬─────────────────────────────┬───────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\BLANK────────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─TEXT(──┴─'description'─┴──)─┘
5──┬──────────────────────────────────────────┬──)───────────────────────5%
│ ┌─\NONE────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
└─JOB(──┴──6─┬────────────┬──job-name─┴─┴──
(1)
─┘
└─user-name/─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 A maximum of 250 repetitions

Purpose
The Create Functional Area (CRTFCNARA) command allows the user to create
functional areas on the system. Functional areas are used by Performance
Tools/400 for reports and graphics. A functional area is a pre-defined list of job
names and/or user names that are to be included in a report or graph.

Examples
Example 1: Creating a Functional Area in the Default Library
CRTFCNARA FCNARA(PERSONNEL)
JOB(MIKE/\N ROSS/\N QPFR\)

This command creates the functional area PERSONNEL with three entries:
Ÿ The user MIKE
Ÿ The user ROSS
Ÿ Any job beginning with QPFR
The functional area is created in the QPFRDATA library.

A-20 Performance Tools V4R2

 CRTGPHFMT

Example 2: Creating a Functional Area in a Specified Library

CRTFCNARA FCNARA('Performance Tools')
LIB(RPFT) JOB(TODD/\N MARTY/\N DEB/QPFRMON)

This command creates the functional area 'Performance Tools' with three entries:
Ÿ The user TODD
Ÿ The user MARTY
Ÿ Any QPFRMON job submitted by DEB
The functional area is created in the RPFT library.

CRTGPHFMT (Create Graph Format) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
55──CRTGPHFMT──GPHFMT(──┼───────────────┼──format-name──)─────────────────5
└─library-name/─┘
(P)
5──┬─────────────────────────────┬──┬──────────────────────────────┬── ────5
│ ┌─\BLANK────────┐ │ │ ┌─\BLANK────────┐ │
└─TEXT(──┴─'description'─┴──)─┘ └─TITLE(──┼─\MBRTEXT──────┼──)─┘
└─'graph-title'─┘
5──┬────────────────────────────────────┬─────────────────────────────────5
│ ┌─\BLANK───────────┐ │
└─SUBTITLE(──┼─\MBRTEXT─────────┼──)─┘
└─'graph-subtitle'─┘
5──┬───────────────────────────┬──┬─────────────────────────────┬─────────5
│ ┌─\SURFACE─┐ │ │ ┌─\ALL──────┐ │
└─GPHTYPE(──┼─\LINE────┼──)─┘ └─DATATYPE(──┼─\FCNARA───┼──)─┘
├─\CBAR────┤ ├─\JOBTYPE──┤
├─\FBAR────┤ ├─\PRIORITY─┤
└─\SCATTER─┘ ├─\IOP──────┤
├─\DISK─────┤
└─\CMNLINE──┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─\NO──┐ │
└─AREAFILL(──┴─\YES─┴──)─┘
5──┬────────────────────────────────────────┬──┬───────────────────┬──────5
│ ┌─\NONE─────────────────┐ │ └─XAXIS(──┤ Xaxis ├─┘
└─REFLINE(──┴─reference-line-number─┴──)─┘
5──┬───────────────────┬──┬──────────────────────────────────────┬────────5
└─YAXIS(──┤ Yaxis ├─┘ │ ┌──
──────────────────┐ │
(1) 6 (2)
└─FCNARA(─── ───(──┤ Fcnara ├──)─┴── ──)─┘

Appendix A. Performance Tools CL Commands A-21

CRTGPHFMT

5──┬────────────────────────────────────────────────────────────────┬─────5
│ ┌──
───────────────────────────────────────────┐ │
(3) 6 (4)
└─JOBTYPE(─── ───(──┬─\ALL───────┬──┬─────────────────┬──)─┴── ──)─┘
├─\ALLINTER──┤ └─┤ More Values ├─┘
├─\ALLBATCH──┤
├─\ALLSYSTEM─┤
├─\ASJ───────┤
├─\BCH───────┤
├─\CA4───────┤
├─\DDM───────┤
├─\EVK───────┤
├─\INT───────┤
├─\MRT───────┤
├─\PCS───────┤
├─\PDJ───────┤
├─\PJ────────┤
├─\PTH───────┤
├─\RDR───────┤
├─\S36───────┤
├─\SBS───────┤
├─\SYS───────┤
├─\WTR───────┤
└─\OTHER─────┘
5──┬────────────────────────────────────────────────────────────┬────────5%
│ ┌──
──────────────┐ │
├─PRIORITY(─── (5)
────6─┤ Priority ├─┴──(6)
────)─────────────────────┤
│ ┌──
─────────────────────────────────────┐ │
(7)
├─IOP(─── ────6─(──┬─\AVG─┬──┬─────────────────┬──)─┴── (8)
────)───┤
│ └─\MAX─┘ └─┤ More Values ├─┘ │
│ ┌──
─────────────────────────────────────┐ │
(9)
├─DISK(─── ────6─(──┬─\AVG─┬──┬─────────────────┬──)─┴─── (10)
────)─┤
│ └─\MAX─┘ └─┤ More Values ├─┘ │
│ ┌──
─────────────┐ │
├─CMNLINE(────(11)
────6─┤ Cmnline ├─┴───
(12)
────)─────────────────────┤
(13)
└─ALLDATA(──── ───┬─────────────────┬────)────────────────────┘
└─┤ More Values ├─┘
More Values:
├──┬────────────────────────────────────────────────────────────┬─────────┤
│ ┌─\DFT──────────┐ │
└─┼─\BLANK────────┼──┬─────────────────────────────────────┬─┘
(14)
└─'description'─┘ │ ┌─7──── ────────────┐ ┌─\─────────┐ │
└─┴─line-type-number─┴──┼───────────┼─┘
└─character─┘
Xaxis:
┌─\TIME─────┐
├──┼─\CPU──────┼──┬──────────────────────────────────────────────────┬──)──┤
├─\TNS──────┤ │ ┌─\DFT───────────┐ ┌─\AUTO────────────────────┐ │
├─\NBRTNS───┤ └─┼─\BLANK─────────┼──┼──────────────────────────┼─┘
├─\RSP──────┤ └─'x-axis-title'─┘ └─start-number──end-number─┘
├─\SYNCIO───┤
├─\NBRSYNC──┤
├─\ASYNCIO──┤
├─\NBRASYNC─┤
├─\TOTDSKIO─┤
└─\NBROSKIO─┘
Yaxis:

A-22 Performance Tools V4R2

 CRTGPHFMT

┌─\CPU───────┐
├──┼─\TNS───────┼──┬──────────────────────────────────────────────────┬──)─
─┤
├─\NBRTNS────┤ │ ┌─\DFT───────────┐ ┌─\AUTO────────────────────┐ │
├─\RSP───────┤ └─┼─\BLANK─────────┼──┼──────────────────────────┼─┘
├─\SYNCIO────┤ └─'y-axis-title'─┘ └─start-number──end-number─┘
├─\NBRSYNC───┤
├─\ASYNCIO───┤
├─\NBRASYNC──┤
├─\TOTDSKIO──┤
├─\NBRDSKIO──┤
├─\CMNIOP────┤
├─\DSKIOP────┤
├─\LWSIOP────┤
├─\MFCIOP────┤
├─\MFDIOP────┤
├─\DSKARM────┤
├─\PCTDSKOCC─┤
├─\CMNLINE───┤
| └─\LGLDBIO───┘
Fcnara:
├──┬─\OTHER───────────────┬──┬─────────────────┬──────────────────────────┤
└─functional-area-name─┘ └─┤ More Values ├─┘
Priority:
├──(──┬─\ALL─────────────────────┬──┬─────────────────┬──)────────────────┤
├─\OTHER───────────────────┤ └─┤ More Values ├─┘
└─lower-limit──upper-limit─┘
Cmnline:
├──(──┬─\MAX─────────────────────┬──┬─────────────────┬──)────────────────┤
└─communications-line-name─┘ └─┤ More Values ├─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 Valid only when DATATYPE(*FCNARA) is specified

2 A maximum of 16 repetitions

3 Valid only when DATATYPE(*JOBTYPE) is specified

4 A maximum of 16 repetitions

5 Valid only when DATATYPE(*PRIORITY) is specified

6 A maximum of 16 repetitions

7 Valid only when *DATATYPE(*IOP) is specified

8 A maximum of 2 repetitions

9 Valid only when DATATYPE(*DISK) is specified.

10 A maximum of 2 repetitions

11 Valid only when DATATYPE(*CMNLINE) is specified.

12 A maximum of 16 repetitions

13 Valid only when DATATYPE(*ALL) is specified.

14 Solid line

Purpose
The Create Graph Format (CRTGPHFMT) command creates a graph format used
to display performance and historical graphs that are created from performance
data members.

Appendix A. Performance Tools CL Commands A-23

CRTGPHPKG

Example
CRTGPHFMT GPHFMT(FORMAT1) TITLE(\MBRTEXT)
DATATYPE(\FCNARA)
FCNARA((ACCOUNTING 'ACCOUNTING')
(SALES 'SALES' 7 #) (OFFICE 'OFFICE' 7 @))

This command creates a graph format named FORMAT1 in the QPFRDATA library.
The member that is presented on the graph using this format supplies the title for
the graph. The graph is a surface graph with no area fill nor a reference line. The
jobs presented on the graph are grouped according to three functional areas:
1. ACCOUNTING
2. SALES
3. OFFICE
The functional area of accounting is represented on the graph with a solid line with
the label, ACCOUNTING. If the format is displayed on a non-graphics work station,
the asterisk (*) symbol is used to graphically represent the functional area of
accounting. The functional areas, sales and office, are formatted in the same
manner as accounting, except SALES is graphically represented with the # symbol
on a non-graphics work station, and OFFICE is graphically represented with the at
@ symbol on a non-graphics work station.

CRTGPHPKG (Create Graph Package) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P)
55──CRTGPHPKG──GPHPKG(──┼───────────────┼──package-name──)─── ─────────────5
├─\CURLIB/──────┤
└─library-name/─┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─\BLANK────────┐ │
└─TEXT(──┴─'description'─┴──)─┘
5──┬───────────────────────────────────┬─────────────────────────────────5%
(1)
│ ┌─\SELECT─── ────────┐ │
│ │ ┌──
─────────────┐ │ │
└─GPHFMT(──┴──6─format-name─┴──(2)
─┴──)─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *SELECT not in batch.

2 A maximum of 25 repetitions

Purpose
The Create Graph Package (CRTGPHPKG) command creates a graph package of
one or more graph formats.

A-24 Performance Tools V4R2

 CRTHSTDTA

Examples
Example 1: Creating a Package Containing Three Formats
CRTGPHPKG GPHPKG(EXAMPLE)
TEXT('THIS IS AN EXAMPLE')
GPHFMT(GPH1 GPH9 GPH12)

This command creates a graph package called EXAMPLE, which contains three
formats, GPH1, GPH9, and GPH12. This package is saved in the default library,
QPFRDATA.

Example 2: Creating a Package Containing Two Formats

CRTGPHPKG GPHPKG(MYLIB/MYPKG)
TEXT('MY PACKAGE') GPHFMT(MYGPH1 MYGPH2)

This command creates a graph package called MYPKG which contains the formats
of MYGPH1 and MYGPH2. MYPKG is saved in library MYLIB.

CRTHSTDTA (Create Historical Data) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──CRTHSTDTA──MBR(──member-name──)──┬───────────────────────────┬── ──────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┼─\CURLIB──────┼──)─┘
└─library-name─┘
5──┬───────────────────────┬──────────────────────────────────────────────5
│ ┌─\NO──┐ │
└─REPLACE(──┴─\YES─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Create Historical Data (CRTHSTDTA) command creates the files required to
create historical graphs. If the files already exist, this command updates the files to
contain the specified data of a member. Historical data is an ongoing summary of
the system containing the members that have been summarized using this
command.

Appendix A. Performance Tools CL Commands A-25

DLTFCNARA

Examples
Example 1: Creating Files in Default Library
CRTHSTDTA MBR(MONDAY)

This command creates files containing historical data that contain the member
named MONDAY.

Example 2: Creating Files in Specified Library

CRTHSTDTA MBR(TUESPM) LIB(MYLIB)

This command creates files containing historical data that contain the member
named TUESPM located in library MYLIB.

DLTFCNARA (Delete Functional Area) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──DLTFCNARA──FCNARA(──functional-area-name──)───────────────────────────5
(P)
5──┬───────────────────────────┬── ───────────────────────────────────────5%
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Delete Functional Area (DLTFCNARA) command allows the user to delete
functional areas from the system. Functional areas are used by Performance
Tools/400 for reports and graphics. A functional area is a pre-defined list of job
names and user names that are included in a report or graph.

Examples
Example 1: Deleting the Functional Area from the Default Library
DLTFCNARA FCNARA(PERSONNEL)

This command deletes the functional area named PERSONNEL from library
QPFRDATA.

Example 2: Deleting the Functional Area from a Specified Library

DLTFCNARA FCNARA('Performance Tools')
LIB(RPFT)

This command deletes the functional area named 'Performance Tools' from library
RPFT.

A-26 Performance Tools V4R2

 DLTGPHPKG

DLTGPHFMT (Delete Graph Format) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P)
55──DLTGPHFMT──GPHFMT(──┼───────────────┼──format-name──)─── ─────────────5%
├─\CURLIB/──────┤
└─library-name/─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Delete Graph Format (DLTGPHFMT) command deletes a specific graph
format. This command also deletes the specified graph format from any packages
that contain it.

Examples
Example 1: Deleting a Graph Format in a Specified Library

DLTGPHFMT GPHFMT(MYLIB/MYFMT)

This command deletes the graph format named MYFMT located in the MYLIB
library.

Example 2: Deleting a Graph Format in the Default Library

DLTGPHFMT GPHFMT(TESTFMT)

This command deletes the graph format named TESTFMT located in the default
library QPFRDATA.

DLTGPHPKG (Delete Graph Package) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
(P)
55──DLTGPHPKG──GPHPKG(──┼───────────────┼──package-name──)─── ────────────5%
├─\CURLIB/──────┤
└─library-name/─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Appendix A. Performance Tools CL Commands A-27

DLTHSTDTA

Purpose
The Delete Graph Package (DLTGPHPKG) command deletes a graph package.

Examples
Example 1: Deleting a Package in a Specified Library
DLTGPHPKG GPHPKG(MYLIB/MYPKG)

This command deletes the graph package MYPKG from library MYLIB.

Example 2: Deleting a Package in the Default Library

DLTGPHPKG GPHPKG(TESTPKG)

This command deletes graph package TESTPKG from default library QPFRDATA.

DLTHSTDTA (Delete Historical Data) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──DLTHSTDTA──┬────────────────────────────────┬─────────────────────────5
(1)
│ ┌─\SELECT─── ────────┐ │
│ │ ┌──
─────────────┐ │ │
└─MBR(──┴──6─member-name─┴──(2)
─┴──)─┘
(P)
5──┬───────────────────────────┬── ────────────────────────────────────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
1 *SELECT not valid in batch.
2 A maximum of 50 repetitions

P All parameters preceding this point can be specified in positional form.

Purpose
The Delete Historical Data (DLTHSTDTA) command deletes historical data for the
specified members from the historical database files.

Example
DLTHSTDTA MBR(TEST1)

This command deletes historical database file member TEST1 from the historical
database files found in the library QPFRDATA. QPFRJOBD is used for the job
description of the job.

A-28 Performance Tools V4R2

 DLTPFRDTA

DLTPEXDTA (Delete Performance Explorer Data) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──DLTPEXDTA──DTAMBR(──data-member-name──)─── ────────────────────────────5
5──┬───────────────────────────────────┬─────────────────────────────────5%
│ ┌─QPEXDATA──────────┐ │
└─DTALIB(──┴─data-library-name─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The DLTPEXDTA (Delete Performance Explorer Data) command deletes data that
was collected by the performance explorer tool and was saved across a set of
physical files in a particular library.

Restriction: The user must have object existence authority for each performance
explorer database file in the specified library.

Example
DLTPEXDTA DTAMBR(STATS3)
DTALIB(TESTLIB)

This command will remove members named STATS3 from the performance
explorer database files in library TESTLIB. These members could have been
created when the user ended a performance explorer data collection session
(ENDPEX CL command) specifying 'SSNID(STATS3) DTAOPT(*LIB)
DTALIB(TESTLIB)'.

DLTPFRDTA (Delete Performance Data)

Format
Job: B,I Pgm: B,I REXX: B,I Exec

Appendix A. Performance Tools CL Commands A-29

DSPACCGRP

55──DLTPFRDTA──┬────────────────────────────────┬─────────────────────────5
(1)
│ ┌─\SELECT─── ────────┐ │
│ │ ┌──
─────────────┐ │ │
└─MBR(──┴──6─member-name─┴──(2)
─┴──)─┘
(P)
5──┬───────────────────────────┬── ────────────────────────────────────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
1 *SELECT not valid in batch.
2 A maximum of 50 repetitions

P All parameters preceding this point can be specified in positional form.

Purpose
The Delete Performance Data (DLTPFRDTA) command deletes performance data
members from the performance database files.

Example
DLTPFRDTA MBR(TEST1)

This command deletes performance data member TEST1 from the performance
database files found in library QPFRDATA. QPFRJOBD is used for the job
description of the job.

DSPACCGRP (Display Access Group) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──DSPACCGRP──┬───────────────┬──┬───────────────────────────────┬───────5
└─JOB(──┤ Job ├─┘ │ ┌─\BLANK─────────┐ │
└─TITLE(──┴─'report-title'─┴──)─┘
5──┬────────────────────────┬──┬──────────────────────────┬───────────────5
│ ┌─\──────┐ │ │ ┌─QAPAGDTA────┐ │
└─OUTPUT(──┼─\PRINT─┼──)─┘ └─MBR(──┴─member-name─┴──)─┘
├─\FILE──┤
└─\BOTH──┘
5──┬───────────────────────────┬──┬─────────────────────┬────────────────5%
│ ┌─QPFRDATA─────┐ │ │ ┌─\YES─┐ │
└─LIB(──┴─library-name─┴──)─┘ └─CLEAR(──┴─\NO──┴──)─┘
Job:

A-30 Performance Tools V4R2

 DSPACCGRP

┌─\─────────────────────────────────────────────────────────────┐
├──┼─\SEC──────────────────────────────────────────────────────────┼──)───┤
├─\INT──────────────────────────────────────────────────────────┤
└───┬───────────────────────────────────────────────────────┬───┘
│ ┌─\ALL─────┐ │
└─┬─────────────────────────────────────┬──┴─job-name─┴─┘
│ ┌─\ALL───────┐ │
└─┬─────────────────┬──┴─user-name/─┴─┘
│ ┌─\ALL────────┐ │
└─┴─job-number/─┴─┘

Purpose
The Display Access Group (DSPACCGRP) command shows, for each selected job,
the process access group (PAG) size, the names and I/O counts for all open files,
and the program data storage used. It also optionally writes this information to a
database file for further analysis by the Analyze Process Access Group
(ANZACCGRP) command.

Jobs can be selected for the report using several criteria, including job type, job
name, user name, and job number. The job and user names may be generic; only
the first characters of the corresponding information for a job must match the speci-
fied value for selection of the job. Only jobs that match all specified criteria are
selected.

Examples
Example 1: Displaying Access Group Information
DSPACCGRP JOB(GL\)

This command selects all active jobs with names starting with GL (since GL* is not
one of the special single values, it is interpreted as a three-part job identifier; since
only one part is given, it is taken to be the job name, and the job number and user
name default to *ALL). Since no OUTPUT value is specified, output is displayed if
there is only one GL job. A report is spooled if there is more than one GL job.

Example 2: Writing Access Group Information to a Database File

DSPACCGRP JOB(\ALL) OUTPUT(\FILE) MBR(SAMPLE1)

This command writes the PAG information for all active jobs to member SAMPLE1
of the database file QPFRDATA/QAPTPAGD for later analysis by the ANZACCGRP
command.

Example 3: Displaying All Access Group Information for a Job

DSPACCGRP JOB(ðð1234/\ALL/\ALL)

This command displays the PAG information for job number 001234. Since the job
number is guaranteed to uniquely identify the job, a report cannot be spooled (no
more than one job can be selected).

Example 4: Displaying Access Group Information for a Job

DSPACCGRP JOB(WILLIAMS/AP\)

This command selects all jobs with a user name of WILLIAMS and a job name
beginning with AP.

Appendix A. Performance Tools CL Commands A-31

DSPHSTGPH

DSPHSTGPH (Display Historical Graph) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
55──DSPHSTGPH──GRAPH(──┼───────────────┼──┬─format-name────┬──)───────────5
(1)
├─\CURLIB/──────┤ └─package-name─── ┘
└─library-name/─┘
(P)
5──┬───────────────────────────┬── ──┬──────────────────────────────┬──────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\SAME─────────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─TITLE(──┼─\BLANK────────┼──)─┘
├─\MBRTEXT──────┤
└─'graph-title'─┘
5──┬──────────────────────────────┬──┬──────────────────────────┬─────────5
│ ┌─\SAME──────┐ │ │ ┌─\────────┐ │
└─SUBTITLE(──┼─\BLANK─────┼──)─┘ └─OUTPUT(──┼─\PRINT───┼──)─┘
├─\MBRTEXT───┤ ├─\PLOT────┤
└─'subtitle'─┘ └─\OUTFILE─┘
5──┬────────────────────────────────┬─────────────────────────────────────5
│ ┌─4214─────────┐ │
(2)
└─PRTDEV(─── ─┼─4234─────────┼──)─┘
├─522X─────────┤
├─\IPDS────────┤
├─\NONGRAPHIC──┤
└─printer-name─┘
5──┬──────────────────────────────────────────────────────┬───────────────5
│ ┌─\PRTDEV──────────────────────────────┐ │
│ │ ┌─\LIBL/────────┐ │ │
(3)
└─OUTQ(─── ─┴─┼───────────────┼──output-queue-name─┴──)─┘
└─library-name/─┘
5──┬────────────────────────────┬──┬─────────────────────────────────┬────5
│ ┌─\PRTDEV─┐ │ │ ┌─1ðð───────────┐ │
(4) (5)
└─PAGELEN(─── ─┼─51──────┼──)─┘ └─PLTSPD(─── ─┴─plotter-speed─┴──)─┘
└─66──────┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─3─────────┐ │
(6)
└─PLTPEN(─── ─┴─pen-width─┴──)─┘
5──┬───────────────────────────────────┬──────────────────────────────────5
│ ┌─1───────────────┐ │
(7)
└─PLTADR(─── ─┴─plotter-address─┴──)─┘
5──┬─────────────────────────────────────────────┬────────────────────────5
│ ┌─LIBL/─────────┐ │
(8)
└─OUTFILE(─── ─┼───────────────┼──file-name──)─┘
├─\CURLIB/──────┤
└─library-name/─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\FIRST──────┐ ┌─\REPLACE─┐ │
└─OUTMBR(──┴─member-name─┴──┼──────────┼──)─┘
└─\ADD─────┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\GPHFMT─┐ │
(9)
└─TYPE(─── ─┴─\GPHPKG─┴──)─┘

A-32 Performance Tools V4R2

 DSPHSTGPH

5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\FIRST─────┐ ┌─\LAST────┐ │
(10)
└─PERIOD(──┼─\SELECT──── ┼──┼──────────┼──)─┘
└─start-date─┘ └─end-date─┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\NO──┐ │
└─CRTHSTDTA(──┴─\YES─┴──)─┘
5──┬───────────────────────────────────────────────┬──────────────────────5
│ ┌─\SAME──────────────────────────┐ │
└─XAXIS(──┼─\AUTO──────────────────────────┼──)─┘
└─starting-number──ending-number─┘
5──┬───────────────────────────────────────────────┬──────────────────────5
│ ┌─\SAME──────────────────────────┐ │
└─YAXIS(──┼─\AUTO──────────────────────────┼──)─┘
└─starting-number──ending-number─┘
5──┬─────────────────────────┬──┬────────────────────────┬────────────────5
│ ┌─\SAME─┐ │ │ ┌─DSPHSTGPH─┐ │
└─AREAFILL(──┼─\YES──┼──)─┘ └─JOB(──┴─job-name──┴──)─┘
└─\NO───┘
5──┬──────────────────────────────────────────────────────────┬──────────5%
│ ┌─\NONE───────────────────────────────────┐ │
│ │ ┌─\LIBL/────────┐ │ │
(11)
└─JOBD(──── ─┴─┼───────────────┼──job-description-name─┴──)─┘
├─\CURLIB/──────┤
└─library-name/─┘
Notes:
1 A package-name is not valid if OUTPUT(*OUTFILE) is specified.

TYPE(*GPHPKG) must be specified for graph packages.

P All parameters preceding this point can be specified in positional form.

2 The PRTDEV parameter is valid only when OUTPUT(*PRINT) is specified.

3 The OUTQ parameter is valid only when OUTPUT(*PRINT) is specified and an

actual printer name, for example, PRT01, PRT02, and so on, is not specified.
4 The PRTDEV parameter is valid only when OUTPUT(*PRINT) is specified.

5 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
6 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
7 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
8 The OUTFILE parameter and the OUTMBR parameter are valid only when

OUTPUT(*OUTFILE) is specified.
9 A package-name is not valid if OUTPUT(*OUTFILE) is specified.

TYPE(*GPHPKG) must be specified for graph packages.

10 *SELECT cannot be used in batch environment.

11 JOBD(*NONE) must be specified if OUTPUT(*) or OUTPUT(*PLOT) is specified.

Purpose
The Display Historical Graph (DSPHSTGPH) command produces a graph from the
historical data created by the Create Historical Data (CRTHSTDTA) command. The
DSPHSTGPH command is intended to give the user a historical perspective of the
system in a graphical representation.

The CRTHSTDTA command summarizes the performance data collected by the

Start Performance Monitor (STRPFRMON) command. The graph format must have

Appendix A. Performance Tools CL Commands A-33

DSPHSTGPH

been defined by the Create Graph Format (CRTGPHFMT) command. The graph
can be directed to a graphics terminal, non-graphics terminal, printer, plotter, and a
graphics data format (GDF) file that can be used by other systems. Historical data
members can be selectively included in the graph.
Note: It is important that the CRTHSTDTA command has been run for each of the
members that the user wants to include in the graph. If the CRTHSTDTA
command has not been run for a member, it is not included in the graph
unless CRTHSTDTA(*YES) is specified.

Examples
Example 1: Displaying a Data File
DSPHSTGPH GRAPH(GRAPHLIB/CPU)

This command shows the historical data file in library QPFRDATA on the user's
screen. It is shown using the graph format CPU in library GRAPHLIB. All of the
historical information in library QPFRDATA is included in the graph.

Example 2: Saving a Graph

DSPHSTGPH GRAPH(GRAPHLIB/CPU) OUTPUT(\OUTFILE)
OUTFILE(USERLIB/USERFILE) OUTMBR(TEST)
JOBD(\LIBL/QPFRJOBD)

This command submits a job to save the graph in a GDF file. The graph is saved in
the file USERLIB/USERFILE/TEST.

Example 3: Printing a Graph

DSPHSTGPH GRAPH(GRAPHLIB/CPU) OUTPUT(\PRINT)
PRTDEV(PRTð3) JOBD(\LIBL/QPFRJOBD)

This command submits a job to print the graph on the system printer named
PRT03.

Example 4: Printing All Graphs in a Package

DSPHSTGPH GRAPH(GRAPHLIB/PACKAGE1)
OUTPUT(\PRINT) PRTDEV(PRTð3) TYPE(\GPHPKG)
JOBD(\LIBL/QPFRJOBD)

This command submits a job to print all of the graphs defined in PACKAGE1 in
GRAPHLIB. The print job is sent to the system printer named PRT03. It uses the
historical data members in QPFRDATA for its information.

Example 5: Displaying a Graph

DSPHSTGPH GRAPH(GRAPHLIB/CPU) OUTPUT(\)
PERIOD(112788 1ðð989)

This command displays a graph of historical information from 27 NOV 1988 to 9

OCT 1989.

Example 6: Selecting Members to be in a Graph

DSPHSTGPH GRAPH(GRAPHLIB/CPU) MBRLIB(MONDAY)
OUTPUT(\) PERIOD(\SELECT \N)

A-34 Performance Tools V4R2

 DSPPFRDTA

This command shows the historical members selection menu for the members in
library MONDAY. The user then selects the members to be shown in the graph.

DSPPFRDTA (Display Performance Data) Command

Format
Job: I Pgm: I REXX: I Exec

55──DSPPFRDTA──┬──────────────────────────┬───────────────────────────────5
(1)
│ ┌─\SELECT─── ──┐ │
└─MBR(──┴─member-name─┴──)─┘
(P)
5──┬───────────────────────────┬── ──┬─────────────────────┬──────────────5%
│ ┌─QPFRDATA─────┐ │ └─PERIOD(──┤ Period ├─┘
└─LIB(──┴─library-name─┴──)─┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┼─\SELECT────┼──┼────────────┼──)──┬──────────────────────────────────┬─┤
─
└─start-time─┘ └─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘
Notes:
1 *SELECT cannot be used in a batch environment.
P All parameters preceding this point can be specified in positional form.

Purpose
The Display Performance Data (DSPPFRDTA) command uses a series of displays
to show the performance data collected by the Start Performance Monitor
(STRPFRMON) command.

Examples
Example 1: Displaying List of Members
DSPPFRDTA

This command uses the Select Performance Member display to list the members
available in the default library QPFRDATA so the user can select a member to
display performance data.

Example 2: Displaying Performance Data

DSPPFRDTA MBR(JUNE1)

This command displays the performance data stored in member JUNE1 located in
library QPFRDATA.

Appendix A. Performance Tools CL Commands A-35

DSPPFRGPH

DSPPFRGPH (Display Performance Graph) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─QPFRDATA/─────┐
55──DSPPFRGPH──GRAPH(──┼───────────────┼──┬─format-name────┬──)───────────5
(1)
├─\CURLIB/──────┤ └─package-name─── ┘
└─library-name/─┘
5──MBR(──member-name──)──┬───────────────────────────┬────────────────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬──────────────────────────────┬── ─────────────────────────────────────5
│ ┌─\SAME─────────┐ │
└─TITLE(──┼─\BLANK────────┼──)─┘
├─\MBRTEXT──────┤
└─'graph-title'─┘
5──┬────────────────────────────────────┬─────────────────────────────────5
│ ┌─\SAME────────────┐ │
└─SUBTITLE(──┼─\BLANK───────────┼──)─┘
├─\MBRTEXT─────────┤
└─'graph-subtitle'─┘
5──┬──────────────────────────┬──┬────────────────────────────────┬───────5
│ ┌─\────────┐ │ │ ┌─4214─────────┐ │
(2)
└─OUTPUT(──┼─\PRINT───┼──)─┘ └─PRTDEV(─── ─┼─4234─────────┼──)─┘
├─\PLOT────┤ ├─522X─────────┤
└─\OUTFILE─┘ ├─\IPDS────────┤
├─\NONGRAPHIC──┤
└─printer-name─┘
5──┬──────────────────────────────────────────────────────┬───────────────5
│ ┌─\PRTDEV──────────────────────────────┐ │
│ │ ┌─\LIBL/────────┐ │ │
(3)
└─OUTQ(─── ─┴─┼───────────────┼──output-queue-name─┴──)─┘
└─library-name/─┘
5──┬────────────────────────────┬──┬─────────────────────────────────┬────5
│ ┌─\PRTDEV─┐ │ │ ┌─1ðð───────────┐ │
(4) (5)
└─PAGELEN(─── ─┼─51──────┼──)─┘ └─PLTSPD(─── ─┴─plotter-speed─┴──)─┘
└─66──────┘
5──┬─────────────────────────────┬────────────────────────────────────────5
│ ┌─3─────────┐ │
(6)
└─PLTPEN(─── ─┴─pen-width─┴──)─┘
5──┬───────────────────────────────────┬──────────────────────────────────5
│ ┌─1───────────────┐ │
(7)
└─PLTADR(─── ─┴─plotter-address─┴──)─┘
5──┬─────────────────────────────────────────────┬────────────────────────5
│ ┌─QPFRDATA/─────┐ │
(8)
└─OUTFILE(─── ─┼───────────────┼──file-name──)─┘
├─\LIBL/────────┤
├─\CURLIB/──────┤
└─library-name/─┘
5──┬─────────────────────────────────────────────┬────────────────────────5
│ ┌─\FIRST──────┐ ┌─\REPLACE─┐ │
(9)
└─OUTMBR(─── ─┴─member-name─┴──┼──────────┼──)─┘
└─\ADD─────┘

A-36 Performance Tools V4R2

 DSPPFRGPH

5──┬──────────────────────────┬───────────────────────────────────────────5
│ ┌─\GPHFMT─┐ │
(10)
└─TYPE(──── ─┴─\GPHPKG─┴──)─┘
5──┬───────────────────────────────────────────────┬──────────────────────5
│ ┌─\SAME──────────────────────────┐ │
└─XAXIS(──┼─\AUTO──────────────────────────┼──)─┘
└─starting-number──ending-number─┘
5──┬───────────────────────────────────────────────┬──────────────────────5
│ ┌─\SAME──────────────────────────┐ │
└─YAXIS(──┼─\AUTO──────────────────────────┼──)─┘
└─starting-number──ending-number─┘
5──┬─────────────────────────┬──┬──────────────────────────────┬──────────5
(11)
│ ┌─\SAME─┐ │ └─PERIOD(──── ─(──┤ Period ├──)─┘
└─AREAFILL(──┼─\YES──┼──)─┘
└─\NO───┘
5──┬───────────────────────────┬──────────────────────────────────────────5
(12)
├─SLTJOB(──── ─┤ Sltjob ├──)─┤
(14)
└─OMTJOB(──── ─┤ Omtjob ├──)─┘
5──┬──────────────────────────────────────┬───────────────────────────────5
│ ┌─\ALL─────────────┐ │
│ │ ┌──
───────────┐ │ │
(16)
├─SLTUSER(──── ─┴──6─user-name─┴───
(17)
─┴──)─┤
│ ┌─\NONE─────────┐ │
│ │ ┌──
───────────┐ │ │
(18)
└─OMTUSER(──── ─┴──6─user-name─┴─┴───(19)
──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(20)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(21)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(22)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴─┴───(23)
──)─┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\ALL──────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(24)
├─SLTSBS(──── ─┴──6─subsystem-name─┴───
(25)
─┴──)─┤
│ ┌─\NONE──────────────┐ │
│ │ ┌──
────────────────┐ │ │
(26)
└─OMTSBS(──── ─┴──6─subsystem-name─┴─┴───(27)
──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL────────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(28)
├─SLTLINE(──── ─┴──6─communications-line-name─┴───
(29)
─┴──)─┤
│ ┌─\NONE────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(30)
└─OMTLINE(──── ─┴──6─communications-line-name─┴─┴───(31)
──)─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\ALL───────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(32)
├─SLTCTL(──── ─┴──6─controller-name─┴───
(33)
─┴──)─┤
│ ┌─\NONE───────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(34)
└─OMTCTL(──── ─┴──6─controller-name─┴─┴───(35)
──)─┘
5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(36)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(37)
─┴──)─┤
│ ┌─\NONE────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(38)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴─┴───(39)
──)─┘

Appendix A. Performance Tools CL Commands A-37

DSPPFRGPH

5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─DSPPFRGPH─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬──────────────────────────────────────────────────────────┬──────────5%
│ ┌─\NONE───────────────────────────────────┐ │
│ │ ┌─\LIBL/────────┐ │ │
(40)
└─JOBD(──── ─┴─┼───────────────┼──job-description-name─┴──)─┘
├─\CURLIB/──────┤
└─library-name/─┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┴─start-time─┴──┼────────────┼──)──┬──────────────────────────────────┬─┤
─
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘
Sltjob:
┌─\ALL─────────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴───
(13)
─┴───────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Omtjob:
┌─\NONE─────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴─┴────
(15)
──────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Notes:
1 A package-name is not valid if OUTPUT(*OUTFILE) is specified.
TYPE(*GPHPKG) must be specified for graph packages.
P All parameters preceding this point can be specified in positional form.

2 The PRTDEV and PAGELEN parameters are valid only when OUTPUT(*PRINT)

is specified.
3 The OUTQ parameter is valid only when OUTPUT(*PRINT) is specified and an

actual printer name, for example, PRT01, PRT02, and so on, is not specified.
4 The PRTDEV and PAGELEN parameters are valid only when OUTPUT(*PRINT)

is specified.
5 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
6 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
7 The PLTSPD, PLTPEN, and PLTADR parameters are valid only when

OUTPUT(*PLOT) is specified.
8 The OUTFILE and OUTMBR parameters are valid only when

OUTPUT(*OUTFILE) is specified.
9 The OUTFILE and OUTMBR parameters are valid only when

OUTPUT(*OUTFILE) is specified.
10 A package-name is not valid if OUTPUT(*OUTFILE) is specified.

TYPE(*GPHPKG) must be specified for graph packages.

11 PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
12 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the

same time.

A-38 Performance Tools V4R2

 DSPPFRGPH

13 A maximum of 50 repetitions
14 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the
same time.
15 A maximum of 50 repetitions
16 The SLTUSER parameter and the OMTUSER parameter cannot be specified at
the same time.
17 A maximum of 50 repetitions
18 The SLTUSER parameter and the OMTUSER parameter cannot be specified at
the same time.
19 A maximum of 50 repetitions
20 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified
at the same time
21 A maximum of 16 repetitions
22 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified
at the same time
23 A maximum of 16 repetitions
24 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the
same time.
25 A maximum of 50 repetitions
26 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the
same time.
27 A maximum of 50 repetitions
28 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the
same time.
29 A maximum of 50 repetitions
30 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the
same time.
31 A maximum of 50 repetitions
32 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
33 A maximum of 50 repetitions
34 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
35 A maximum of 50 repetitions
36 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
37 A maximum of 50 repetitions
38 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
39 A maximum of 50 repetitions
40 The JOBD parameter is valid only when OUTPUT(*) or OUTPUT(*PLOT) is not
specified.

Purpose
The Display Performance Graph (DSPPFRGPH) command produces a graph from
the performance data collected by the Start Performance Monitor (STRPFRMON)
command.

The graph format must have been defined on the Create Graph Format
(CRTGPHFMT) command. The graph can be sent as output to a graphics terminal,

Appendix A. Performance Tools CL Commands A-39

DSPPFRGPH

a non-graphics terminal, a printer, a plotter, or to a graphics data format (GDF) file

that can be used by other systems. Jobs can be selectively included or omitted
from the graph.

Examples
Example 1: Displaying Performance Data Files
DSPPFRGPH GRAPH(GRAPHLIB/CPU)
MBR(QPFRDATA/JUN1) TITLE(\MBRTEXT)

This command displays the performance data files in library QPFRDATA member
JUN1. It is displayed as specified by graph format CPU in library GRAPHLIB. The
report title is taken from the text of the member.

Example 2: Saving a Graph

DSPPFRGPH GRAPH(GRAPHLIB/CPU)
MBR(JUN1) OUTPUT(\OUTFILE)
OUTFILE(USERLIB/USERFILE) OUTMBR(TEST)
JOBD(\LIBL/QPFRJOBD)

This command submits a job to save the graph of performance data from file
member JUN1, which is in library QPFRDATA, in a GDF file. The graph is saved in
file USERLIB/USERFILE/TEST (file member TEST in file USRFILE, which is in
library USRLIB).

Example 3: Printing a Graph

DSPPFRGPH GRAPH(GRAPHLIB/CPU)
MBR(JUN1) OUTPUT(\PRINT)
PRTDEV(PRTð3) JOBD(\LIBL/QPFRJOBD)

This command submits a job to print the graph of performance data from file
member JUN1, which is in library QPFRDATA, on the system printer named
PRT03.

Example 4: Printing All Graphs

DSPPFRGPH GRAPH(GRAPHLIB/PACKAGE1)
MBR(JUN1) OUTPUT(\PRINT)
PRTDEV(PRTð3) JOBD(\LIBL/QPFRJOBD)
TYPE(\GPHPKG)

This command submits a job to print all of the graphs defined in PACKAGE1 in
GRAPHLIB. The print job is sent to system printer PRT03. It's data source is in
performance data member JUN1 in library QPFRDATA.

Example 5: Displaying a Graph

DSPPFRGPH GRAPH(GRAPHLIB/CPU)
MBR(JUN1) OUTPUT(\)
PERIOD((233ð)(ð13ð))

This command displays a graph of the data collected from 11:30 PM on the first
day of collection through 1:30 AM on the last day of collection. However, if data
collection started and ended on the same day, an error message is printed,
because the specified ending date and time precedes the specified starting date
and time.

A-40 Performance Tools V4R2

 ENDJOBTRC

Example 6: Displaying a Graph

DSPPFRGPH GRAPH(GRAPHLIB/CPU)
MBR(JUN1) OUTPUT(\)
PERIOD((233ð)(ð13ð)) SLTUSER(D46\)

This command displays a graph of the performance data collected for all the jobs
whose user ID starts with D46 from 11:30 PM on the first day of collection through
1:30 AM on the last day of collection. However, if data collection started and ended
on the same day, an error message is printed, because the specified ending date
and time precedes the specified starting date and time.

ENDJOBTRC (End Job Trace) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──ENDJOBTRC──┬──────────────────────────┬───────────────────────────────5
│ ┌─QAJOBTRC────┐ │
└─MBR(──┴─member-name─┴──)─┘
5──┬───────────────────────────┬──┬───────────────────────────┬───────────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\NONE────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─RPTTYPE(──┼─\DETAIL──┼──)─┘
├─\SUMMARY─┤
└─\BOTH────┘
(P)
5──┬───────────────────────────────┬── ────────────────────────────────────5
│ ┌─\BLANK─────────┐ │
└─TITLE(──┴─'report-title'─┴──)─┘
5──┬─────────────────────────────────┬────────────────────────────────────5
│ ┌─\FIRST──────────┐ │
└─STRSEQ(──┴─sequence-number─┴──)─┘
5──┬─────────────────────────────────┬────────────────────────────────────5
│ ┌─\LAST───────────┐ │
└─ENDSEQ(──┴─sequence-number─┴──)─┘
5──┬──────────────────────────────┬──┬──────────────────────────────┬─────5
│ ┌─QT3REQIO─────┐ │ │ ┌─QWSGET───────┐ │
└─ENDTNS(──┴─program-name─┴──)─┘ └─STRTNS(──┴─program-name─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─ENDJOBTRC─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Appendix A. Performance Tools CL Commands A-41

ENDPEX

Purpose
The End Job Trace (ENDJOBTRC) command turns off the OS/400 system job
tracing function that was started by the Start Job Trace (STRJOBTRC) command,
saves all collected trace records in a database file, and optionally produces reports
used to analyze the data from a performance viewpoint (the Print Job Trace
(PRTJOBTRC) command may subsequently be used to produce reports from the
source data).

Examples
Example 1: Stopping Job Tracing
ENDJOBTRC

This command stops tracing and saves the created trace records in
QPFRDATA/QAPTTRCJ, member QAJOBTRC. No reports are produced.

Example 2: Producing a Report of Job Records

ENDJOBTRC LIB(MYLIB) RPTTYPE(\DETAIL)

This command stops job tracing, saves the created trace records in member
QAJOBTRC of MYLIB/QAPTTRCJ, and produces a detail report. The use of a
private library allows several users to trace jobs at the same time.

ENDPEX (End Performance Explorer) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──ENDPEX──SSNID(──session-identifier──)─── ──────────────────────────────5
5──┬──────────────────────────┬──┬─────────────────────────┬──────────────5
│ ┌─\END─────┐ │ │ ┌─\LIB──┐ │
(1)
└─OPTION(──┴─\SUSPEND─┴──)─┘ └─DTAOPT(─── ─┼─\FILE─┼──)─┘
└─\DLT──┘
5──┬──────────────────────────────────────┬───────────────────────────────5
│ ┌─QPEXDATA──────────┐ │
(2, 3)
└─DTALIB(──── ─┴─data-library-name─┴──)─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
│ ┌─\SSNID───────────┐ │
(4, 5)
└─DTAMBR(──── ─┴─data-member-name─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬──────────5
│ ┌─QPEXDATA──────────┐ ┌─\SSNID─────────┐ │
(6, 7)
└─DTAFILE(──── ─┴─data-library-name─┴──┴─data-file-name─┴──)─┘
5──┬────────────────────────┬──┬──────────────────────────┬──────────────5%
│ ┌─\NO──┐ │ │ ┌─\BLANK────────┐ │
(8)
└─RPLDTA(─── ─┴─\YES─┴──)─┘ └─TEXT(──┴─'description'─┴─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 Only valid when OPTION(*END) is specified.

2 Only valid when OPTION(*END) is specified.

3 Only valid when DTAOPT(*LIB) is specified.

4 Only valid when OPTION(*END) is specified.

A-42 Performance Tools V4R2

 PRTACTRPT

5 Only valid when DTAOPT(*LIB) is specified.

6 Only valid when OPTION(*END) is specified.
7 Only valid when DTAOPT(*FILE) is specified.
8 Only valid when OPTION(*END) is specified.

Purpose
The End Performance Explorer (ENDPEX) command instructs the performance
explorer tool to stop collecting data. The command expects a session name to
accompany the request which identifies which instance of the performance explorer
session to end.

The user can either end the data collection session or suspend the data collection
session. If the user chooses to end the session, the collected data is put into a
single physical file or multiple data base files, or it is deleted, based on the value
specified for the DTAOPT parameter.

If the user chooses to suspend the collection of performance data, the session
remains active. To resume data collection for a suspended session, the user can
specify OPTION(*RESUME) on a subsequent call of the STRPEX (Start Perform-
ance Explorer) command.

Examples
Example 1: End a Session and Save the Data
ENDPEX SSNID(TEST3) OPTION(\END)
DTAOPT(\LIB) DTAMBR(SYS1DATA)

This command ends the performance explorer session named TEST3 and saves
the data in a set of database files in library QPEXDATA. The member name to be
used for each file is SYS1DATA.

Example 2: End a Session and Delete the Data

ENDPEX SSNID(TESTRUN)
OPTION(\END) DTAOPT(\DLT)

This command ends the performance explorer session named TESTRUN and
deletes the collected performance data.

PRTACTRPT (Print Activity Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTACTRPT──┬──────────────────────────┬───────────────────────────────5
│ ┌─QAITMON─────┐ │
└─MBR(──┴─member-name─┴──)─┘
(P)
5──┬───────────────────────────┬──┬───────────────────────────────┬── ─────5
│ ┌─QPFRDATA─────┐ │ └─TITLE(──┬─\BLANK─────────┬──)─┘
└─LIB(──┴─library-name─┴──)─┘ └─'report-title'─┘

Appendix A. Performance Tools CL Commands A-43

PRTACTRPT

5──┬───────────────────────────┬──┬───────────────────────┬───────────────5
(1)
│ ┌─\SUMMARY─┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─RPTTYPE(──┼─\DETAIL──┼──)─┘
└─\ALL─────┘
5──┬─────────────────────────┬──┬─────────────────────────────────┬───────5
│ ┌─\CPU───────┐ │ │ ┌─1ð─────────────┐ │
└─SEQ(──┼─\JOBTASK───┼──)─┘ └─NBRJOBS(──┼─\ALL───────────┼──)─┘
├─\USER──────┤ └─number-of-jobs─┘
├─\PTY───────┤
├─\TOTALIO───┤
├─\SYNCIO────┤
├─\ASYNCIO───┤
├─\FAULT─────┤
├─\SDBREAD───┤
├─\SDBWRITE──┤
├─\SNDBREAD──┤
├─\SNDBWRITE─┤
├─\ADBREAD───┤
├─\ADBWRITE──┤
├─\ANDBREAD──┤
└─\ANDBWRITE─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTACTRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┴─start-time─┴──┼────────────┼──)──┬──────────────────────────────────┬─┤
─
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) specified to maintain its position in the param-
eter value sequence.

Purpose
The Print Activity Report (PRTACTRPT) command generates reports based on the
data collected by the Work with System Activity (WRKSYSACT) command.

Examples
Example 1: Generating a Summary Report
PRTACTRPT

This command submits a batch job that generates a summary activity report using
the performance data found in the default member QAITMON located in the default
library QPFRDATA. The report covers the entire measurement period, and the title
of the report is left blank.

Example 2: Generating a Summary and Detailed Activity Report

A-44 Performance Tools V4R2

 PRTCPTRPT

PRTACTRPT MBR(JUNEð1)
TITLE('Activity Report for June 1st')
RPTTYPE(\ALL) SEQ(\CPU)

This command submits a batch job that generates both a summary and a detailed
activity report. The performance data comes from member JUNE01 located in the
default library QPFRDATA. The report covers the entire measurement period, and
the title of the report is 'Activity Report for June 1st.' The detailed activity report lists
ten entries in descending order according to CPU utilization for each interval.

PRTCPTRPT (Print Component Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTCPTRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬───────────────────────┬─────────5
(1)
│ ┌─\MBR───────────┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─TITLE(──┴─'report-title'─┴──)─┘
5──┬───────────────────────────────────────────────────────────┬──────────5
│ ┌─\ALL────────────────────────────────────┐ │
│ │ ┌──
───────────────────────────────────┐ │ │
(2)
├─SLTJOB(─── ─┴──6─┬─job-name──────────────────────┬─┴──
(3)
─┴──)─┤
│ ├─user-name/job name────────────┤ │
│ └─job-number/user-name/job-name─┘ │
│ ┌─\NONE─────────────────────────────────┐ │
│ │ ┌──
───────────────────────────────────┐ │ │
(4)
└─OMTJOB(─── ─┴──6─┬─job-name──────────────────────┬─┴─┴── (5)
──)─┘
├─user-name/job-name────────────┤
└─job-number/user-name/job-name─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
│ ┌─\ALL────────────┐ │
│ │ ┌──
───────────┐ │ │
(6)
├─SLTUSRID(─── ─┴──6─user-name─┴──
(7)
─┴──)─┤
│ ┌─\NONE─────────┐ │
│ │ ┌──
───────────┐ │ │
(8)
└─OMTUSRID(─── ─┴──6─user-name─┴─┴── (9)
──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(10)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(11)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(12)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴─┴───(13)
──)─┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\ALL──────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(14)
├─SLTSBS(──── ─┴──6─subsystem-name─┴───
(15)
─┴──)─┤
│ ┌─\NONE──────────────┐ │
│ │ ┌──
────────────────┐ │ │
(16)
└─OMTSBS(──── ─┴──6─subsystem-name─┴─┴───(17)
──)─┘

Appendix A. Performance Tools CL Commands A-45

PRTCPTRPT

5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL────────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(18)
├─SLTLINE(──── ─┴──6─communications-line-name─┴───
(19)
─┴──)─┤
│ ┌─\NONE────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(20)
└─OMTLINE(──── ─┴──6─communications-line-name─┴─┴───(21)
──)─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\ALL───────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(22)
├─SLTCTL(──── ─┴──6─controller-name─┴───
(23)
─┴──)─┤
│ ┌─\NONE───────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(24)
└─OMTCTL(──── ─┴──6─controller-name─┴─┴───(25)
──)─┘
5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(26)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(27)
─┴──)─┤
│ ┌─\NONE────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(28)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴─┴───(29)
──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTCPTRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┼─\SELECT────┼──┴─start-date─┴──)──┬──────────────────────────────────┬─┤
─
└─start-time─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
2 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the

same time.
3 A maximum of 50 repetitions

4 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the

same time.
5 A maximum of 50 repetitions

6 The SLTUSRID parameter and the OMTUSRID parameter cannot be specified at

the same time.

7 A maximum of 50 repetitions

8 The SLTUSRID parameter and the OMTUSRID parameter cannot be specified at

the same time.

9 A maximum of 50 repetitions

10 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified

at the same time

A-46 Performance Tools V4R2

 PRTCPTRPT

11 A maximum of 16 repetitions.
12 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified
at the same time
13 A maximum of 16 repetitions.
14 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the
same time.
15 A maximum of 50 repetitions
16 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the
same time.
17 A maximum of 50 repetitions
18 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the
same time.
19 A maximum of 50 repetitions
20 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the
same time.
21 A maximum of 50 repetitions
22 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
23 A maximum of 50 repetitions
24 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
25 A maximum of 50 repetitions
26 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
27 A maximum of 50 repetitions
28 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
29 A maximum of 50 repetitions

Purpose
The Print Component Report (PRTCPTRPT) command produces a report that
expands on the detail for each component of system performance shown on the
System Report. This report is produced from the performance data collected by the
Start Performance Monitor (STRPFRMON) command and shows the data by job,
user, pool, disk, local work station, and exception. This report is written to the
printer file QPPTCPTR. Jobs may be selectively included in the report or excluded
from the report based on a variety of job details and interval times.

Examples
Example 1: Printing a Component Report
PRTCPTRPT MBR(APRIL18)

This command prints a complete component report for the performance data
member APRIL18 in library QPFRDATA. The report title is the same as the text in
the member.

Example 2: Printing a Report With a Title

PRTCPTRPT MBR(NOV1) PERIOD(\SELECT)
TITLE('Intervals with Highest Response Times')

Appendix A. Performance Tools CL Commands A-47

PRTJOBRPT

This command prints a component report for the data member NOV1 in library
QPFRDATA. The user is presented with the interval-selection display, which allows
sorting of the intervals according to various criteria and selection of only certain
intervals to be included in the report. The title of the report is "Intervals with Highest
Response Times".

PRTJOBRPT (Print Job Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTJOBRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬─────────────────────────┬───────5
(1)
│ ┌─\MBRTXT────────┐ │ │ ┌─\SELECT─── ─┐ │
└─TITLE(──┼─\BLANK─────────┼──)─┘ └─PERIOD(──┴─┤ Period ├─┴─┘
└─'report-title'─┘
5──┬──────────────────────────┬───────────────────────────────────────────5
(2)
├─SLTJOB(─── ─┤ Sltjob ├──)─┤
(4)
└─OMTJOB(─── ─┤ Omtjob ├──)─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
│ ┌─\ALL────────────┐ │
│ │ ┌──
───────────┐ │ │
(6)
├─SLTUSRID(─── ─┴──6─user-name─┴──
(7)
─┴──)─┤
│ ┌─\NONE───────────┐ │
│ │ ┌──
───────────┐ │ │
(8)
└─OMTUSRID(─── ─┴──6─user-name─┴──
(9)
─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(10)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(11)
─┴──)─┤
│ ┌─\NONE──────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(12)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(13)
─┴──)─┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\ALL──────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(14)
├─SLTSBS(──── ─┴──6─subsystem-name─┴───
(15)
─┴──)─┤
│ ┌─\NONE─────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(16)
└─OMTSBS(──── ─┴──6─subsystem-name─┴───
(17)
─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL────────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(18)
├─SLTLINE(──── ─┴──6─communications-line-name─┴───
(19)
─┴──)─┤
│ ┌─\NONE───────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(20)
└─OMTLINE(──── ─┴──6─communications-line-name─┴───
(21)
─┴──)─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\ALL───────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(22)
├─SLTCTL(──── ─┴──6─controller-name─┴───
(23)
─┴──)─┤
│ ┌─\NONE──────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(24)
└─OMTCTL(──── ─┴──6─controller-name─┴───
(25)
─┴──)─┘

A-48 Performance Tools V4R2

 PRTJOBRPT

5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(26)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(27)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(28)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴───
(29)
─┴──)─┘
5──┬─────────────────────────┬──┬────────────────────────┬────────────────5
│ ┌─\YES─┐ │ │ ┌─PRTJOBRPT─┐ │
└─OMTSYSTSK(──┴─\NO──┴──)─┘ └─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:
┌─\FIRST─────┐ ┌─\FIRST─────┐
├──┴─start-time─┴──┼────────────┼──)──┬──────────────────────────────────┬─┤
─
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
└─end-date─┘
Sltjob:
┌─\ALL────────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(3)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Omtjob:
┌─\NONE───────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(5)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *SELECT is valid only in the interactive environment.

2 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

3 A maximum of 50 repetitions

4 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

5 A maximum of 50 repetitions

6 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-

sive.
7 A maximum of 50 repetitions

8 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-

sive.
9 A maximum of 50 repetitions

10 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-

sive.
11 A maximum of 16 repetitions

12 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-

sive.
13 A maximum of 16 repetitions

Appendix A. Performance Tools CL Commands A-49

PRTJOBRPT

14 The SLTSBS parameter and the OMTSBS parameter are mutually exclusive.
15 A maximum of 50 repetitions
16 The SLTSBS parameter and the OMTSBS parameter are mutually exclusive.
17 A maximum of 50 repetitions
18 The SLTLINE parameter and the OMTLINE parameter are mutually exclusive.
19 A maximum of 50 repetitions
20 The SLTLINE parameter and the OMTLINE parameter are mutually exclusive.
21 A maximum of 50 repetitions
22 The SLTCTL parameter and the OMTCTL parameter are mutually exclusive.
23 A maximum of 50 repetitions
24 The SLTCTL parameter and the OMTCTL parameter are mutually exclusive.
25 A maximum of 50 repetitions
26 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually
exclusive.
27 A maximum of 50 repetitions
28 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually
exclusive.
29 A maximum of 50 repetitions

Purpose
The Print Job Report (PRTJOBRPT) command produces a job-oriented report from
the performance data collected by the Start Performance Monitor (STRPFRMON)
command. The report, which is written to the printer file QPPTITVJ, shows job infor-
mation by interval. Jobs are selected for inclusion in, or exclusion from, the report
based on a variety of job details and interval times.

Examples
Example 1: Submitting a Batch Job
PRTJOBRPT MBR(DTAð71588A)

This command submits a batch job to print a report on all jobs in all intervals in the
member DTA071588A of the performance data files in library QPFRDATA. The
report title is taken from the text of that member.

Example 2: Selecting Intervals to Include in Report

PRTJOBRPT MBR(DTAð71588A) PERIOD(\SELECT)

This command submits a job to print a report from the same data, but first shows a
screen where a user interactively selects which intervals to include.

Example 3: Reporting on a Specific Time Period

PRTJOBRPT MBR(DTAð71588A) PERIOD((233ð)(ð13ð))

This command submits a job to print a report on the data collected from 11:30 PM
on the first day of collection through 1:30 AM on the last day of collection.
However, if data collection started and ended on the same day, an error message
is printed instead, because the specified ending date and time is before the speci-
fied starting date and time.

Example 4: Printing a Report Interactively

A-50 Performance Tools V4R2

 PRTJOBTRC

PRTJOBRPT MBR(DTAð71588A) SLTUSRID(D46\)

JOBD(\NONE)

This command interactively prints a report for all jobs with a user ID starting with
D46.

Example 5: Printing a Report Interactively

PRTJOBRPT MBR(DTAð71588A) SLTJOB(D46\/\N)
JOBD(\NONE)

This command performs the same function as the previous example.

PRTJOBTRC (Print Job Trace) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTJOBTRC──┬──────────────────────────┬───────────────────────────────5
│ ┌─QAJOBTRC────┐ │
└─MBR(──┴─member-name─┴──)─┘
5──┬───────────────────────────┬──┬───────────────────────────┬───────────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\BOTH────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─RPTTYPE(──┼─\DETAIL──┼──)─┘
└─\SUMMARY─┘
(P)
5──┬───────────────────────────────┬── ────────────────────────────────────5
│ ┌─\BLANK─────────┐ │
└─TITLE(──┴─'report-title'─┴──)─┘
5──┬─────────────────────────────────┬────────────────────────────────────5
│ ┌─\FIRST──────────┐ │
└─STRSEQ(──┴─sequence-number─┴──)─┘
5──┬─────────────────────────────────┬────────────────────────────────────5
│ ┌─\LAST───────────┐ │
└─ENDSEQ(──┴─sequence-number─┴──)─┘
5──┬──────────────────────────────┬──┬──────────────────────────────┬─────5
│ ┌─QT3REQIO─────┐ │ │ ┌─QWSGET───────┐ │
└─ENDTNS(──┴─program-name─┴──)─┘ └─STRTNS(──┴─program-name─┴──)─┘
5──┬───────────────────────────┬──┬────────────────────────┬──────────────5
│ ┌─\CUR───────┐ │ │ ┌─PRTJOBTRC─┐ │
└─MODEL(──┴─model-code─┴──)─┘ └─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Appendix A. Performance Tools CL Commands A-51

PRTLCKRPT

Purpose
The Print Job Trace (PRTJOBTRC) command produces performance-oriented
reports used to analyze job trace data collected with the Start Job Trace
(STRJOBTRC) and End Job Trace (ENDJOBTRC) commands.

Example
PRTJOBTRC LIB(MYLIB) RPTTYPE(\DETAIL)

This command produces a detail report using data saved in member QAJOBTRC in
library MYLIB/QAPTTRCJ.

PRTLCKRPT (Print Lock Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTLCKRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬──┬───────────────────────┬── ─────────5
│ ┌─\MBRTXT────────┐ │ │ ┌─\SUM─┐ │
└─TITLE(──┼─\BLANK─────────┼──)─┘ └─RPTTYPE(──┼─\TOD─┼──)─┘
└─'report-title'─┘ ├─\RQS─┤
├─\HLD─┤
├─\OBJ─┤
└─\ALL─┘
5──┬─────────────────────┬────────────────────────────────────────────────5
│ ┌─\YES─┐ │
└─FIRST(──┴─\NO──┴──)─┘
5──┬────────────────────────────────────────────┬─────────────────────────5
│ ┌─\FIRST─────┐ ┌─\LAST────┐ │
(1)
└─PERIOD(─── ─┴─start-time─┴──┼──────────┼──)─┘
└─end-time─┘
5──┬─────────────────────────────────────────┬────────────────────────────5
│ ┌─5ðð────────────────────┐ │
└─MINWAIT(──┴─number-of-milliseconds─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTLCKRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *PERIOD contains two lists of one element each. *N must be specified in place

of any omitted element that precedes the value(s) to be specified to maintain its
position in the parameter value sequence.

A-52 Performance Tools V4R2

 PRTPEXRPT

Purpose
The Print Lock Report (PRTLCKRPT) command produces a report that shows lock
and seize conflicts that occur during system operation. This report is produced from
the resource management trace data collected by the Start Performance Monitor
(STRPFRMON) command and formatted by the Print Transaction Report
(PRTTNSRPT) command. This information can be used to determine whether jobs
are being delayed during processing because of unsatisfied lock requests or
internal machine seizes; these conditions are also known as waits.

The input to this command is the output from a previous run of the commands
STRPFRMON TRACE(*ALL) and PRTTNSRPT. The output from this command is
written to the printer file QPPTLCK.

Examples
Example 1: Producing a Summary Report
PRTLCKRPT MBR(RESTRC)

This command produces a summary report from the performance data saved in
member RESTRC of QPFRDATA/QAPMDMPT from a prior run of the Start Per-
formance Monitor (STRPFRMON) and Print Transaction Report (PRTTNSRPT)
commands.

Example 2: Including a Detail Listing Sorted By Time

PRTLCKRPT MBR(RESTRC) RPTTYPE(\TOD)

This command produces the same report as the previous example, except that it
includes a detail listing sorted by the time in which the lock/seize conflicts occurred.

PRTPEXRPT (Print Performance Explorer Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──PRTPEXRPT──MBR(──member-name──)─── ─┬───────────────────────────┬──────5
│ ┌─QPEXDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
5──┬────────────────────────┬──┬──────────────────────────┬───────────────5
| │ ┌─\STATS───┐ │ │ ┌─\PRINT───┐ │
| └─TYPE(──┼─\TRACE───┼──)─┘ └─OUTPUT(──┴─\OUTFILE─┴──)─┘
├─\PROFILE─┤
| └─\BASIC───┘
5──┬────────────────────────────────────────────────────┬─────────────────5
| │ ┌─\LIBL/────────┐ │
| └─OUTFILE(──┼───────────────┼──database-file-name──)─┘
| ├─\CURLIB/──────┤
| └─library-name/─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
| │ ┌─\FIRST──────┐ ┌─\REPLACE─┐ │
| └─OUTMBR(──┴─member-name─┴──┴─\ADD─────┴──)─┘

Appendix A. Performance Tools CL Commands A-53

 PRTPEXRPT

5──┬─────────────────────────────────────────────────────────────────┬────5
| │ ┌─\NONE─────────────┐ │
| │ ┌─\TASK──────┐ ┌─\NO──┐ │ ┌── ─────────────┐ │ │
| (1)
└─TRACEOPT(─── ─┴─\TIMESTAMP─┴──┴─\YES─┴──┴──6┬─\PGM──────┬┴──
(2)
─┴──)─┘
| ├─\LICPGM───┤
| ├─\ASM──────┤
| ├─\BASE─────┤
| ├─\DISK─────┤
| ├─\DSKSVR───┤
| ├─\FAULT────┤
| ├─\JOB──────┤
| ├─\LOCK─────┤
| ├─\SAR──────┤
| ├─\MIBRKT───┤
| ├─\LICBRKT──┤
| ├─\DASD─────┤
| ├─\DASDSRVR─┤
| ├─\PAGEFLT──┤
| ├─\RMPR─────┤
| └─\RMSZ─────┘
5──┬──────────────────────────────────────┬───────────────────────────────5
| │ ┌─\ALL────────────────┐ │
| │ │ ┌──
─────────────────┐ │ │
| └─TRCTYPE(──┴──6─┬─\CALLRTN──┬──
(3)
─┴─┴──)─┘
| ├─\BASIC────┤
| ├─\DSKIO1───┤
| ├─\DSKIO2───┤
| ├─\DSKSVR───┤
| ├─\DSKSTG───┤
| ├─\VRTADR───┤
| ├─\PGMACT───┤
| ├─\FILEOPEN─┤
| ├─\PRFDTA───┤
| └─\TASKSWT──┘
5──┬──────────────────────────────────────────────────────┬───────────────5
| │ ┌─\AVAIL─────┐ │
(4, 5)
| └─PERIOD(──── ─(──┴─start-time─┴──┤ PERIOD details ├──)─┘
5──┬─────────────────────────────────────────────────────────────┬────────5
| │ ┌─\ALL──────────────────────────────────────┐ │
| │ │ ┌──
─────────────────────────────────────┐ │ │
| │ │ │ ┌─\ALL/───────┐ │ │ │
| (6)
├─SLTJOB(─── ─┴──6─┴─job-number/─┴──┤ SLTJOB details ├─┴──
(7)
─┴──)─┤
| │ ┌─\NONE─────────────────────────────────────┐ │
| │ │ ┌──
─────────────────────────────────────┐ │ │
| │ │ │ ┌─\ALL/───────┐ │ │ │
| (8)
└─OMTJOB(─── ─┴──6─┴─job-number/─┴──┤ OMTJOB details ├─┴──
(9)
─┴──)─┘
5──┬────────────────────────────────────────────────────┬─────────────────5
│ ┌─\CPU────────────┐ ┌─\PROGRAM─┐ │
(10)
└─STATSOPT(──── ─┼─\PGMNAME────────┼──┼─\BLANK───┼──)─┘
├─\INVCNT─────────┤ └─\MODULE──┘
├─\DBSYNCIO───────┤
├─\DBASYNCIO──────┤
├─\NDBSYNCIO──────┤
├─\NDBASYNCIO─────┤
├─\MICALLS────────┤
├─\MIINST─────────┤
├─\CUMLCPU────────┤
├─\CUMLDBSYNCIO───┤
├─\CUMLDBASYNCIO──┤
├─\CUMLNDBSYNCIO──┤
└─\CUMLNDBASYNCIO─┘

A-54 Performance Tools V4R2

 PRTPEXRPT

5──┬───────────────────────────────────────────┬──────────────────────────5
(11)
└─PROFILEOPT(──── ─┤ PROFILEOPT details ├──)─┘
5──┬────────────────────────────┬────────────────────────────────────────5%
│ ┌─\DESCENDING─┐ │
└─ORDER(──┴─\ASCENDING──┴──)─┘
PERIOD details:
┌─\BEGIN─────┐ ┌─\AVAIL───┐ ┌─\END─────┐
├──┼─\CURRENT───┼──)──(──┴─end-time─┴──┼─\CURRENT─┼──)────────────────────┤
└─start-date─┘ └─end-date─┘
SLTJOB details:
┌─\ALL/───────────────┐
├──┼─generic\-user-name/─┼──┬─job-name──────────┬─────────────────────────┤
└─user-name/──────────┘ └─generic\-job-name─┘
OMTJOB details:
┌─\ALL/───────────────┐
├──┼─generic\-user-name/─┼──┬─job-name──────────┬─────────────────────────┤
└─user-name/──────────┘ └─generic\-job-name─┘
PROFILEOPT details:
┌─\SAMPLECOUNT─┐ ┌─\PROGRAM───┐ ┌─ð─────────────────┐
├──┴─\ADDRESS─────┴──┼─\BLANK─────┼──┴─filter-percentage─┴────────────────┤
├─\STATEMENT─┤
├─\PROCEDURE─┤
└─\MODULE────┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 The TRACEOPT parameter is valid only if TYPE(*TRACE) specified.

| 2 A maximum of 10 values can be specified.

| 3 A maximum of 10 values can be specified.

| 4 This parameter is valid only if TYPE(*TRACE) is specified.

| 5 *PERIOD contains two lists of two elements each. *N must be specified for any

| element that precedes the value or values to be specified, to maintain its position
| in the parameter value sequence.
| 6 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

| 7 A maximum of 10 repetitions.

| 8 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

| 9 A maximum of 10 repetitions.

10 The STATSOPT parameter is valid only if TYPE(*TRACE) specified.

11 The PROFILEOPT parameter is valid only if TYPE(*PROFILE) specified.

Purpose
The Print Performance Explorer Report (PRTPEXRPT) command prints a formatted
listing of the data that was collected by the performance explorer and saved across
a set of physical files in a particular library.

Restriction: The user must have read authority for each performance explorer
database file in the specified library.

Appendix A. Performance Tools CL Commands A-55

PRTPOLRPT

Examples
Example 1: Statistics Report
PRTPEXRPT MBR(SAMPLE) LIBRARY(SAMPLELIB)
TYPE(\STATS) STATSOPT(\INVCNT \MODULE)

In this example, a statistics type report is generated based on data members

named SAMPLE in library SAMPLELIB. The data is arranged in descending order
based on invocation counts and is summarized at the module level.

Example 2: Profile Report

PRTPEXRPT MBR(SAMPLE2) TYPE(\PROFILE)
PROFILEOPT(\SAMPLECOUNT \PROGRAM)
ORDER(\DESCENDING)

In this example, a profile type report is generated based on data members named
SAMPLE2 in the default library, QPEXDATA. The data is arranged in descending
order based on the sample count and is summarized at the program level.

PRTPOLRPT (Print Pool Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTPOLRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬───────────────────────┬─────────5
(1)
│ ┌─\MBRTXT────────┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─TITLE(──┼─\BLANK─────────┼──)─┘
└─'report-title'─┘
5──┬──────────────────────────┬───────────────────────────────────────────5
(2)
├─SLTJOB(─── ─┤ Sltjob ├──)─┤
(4)
└─OMTJOB(─── ─┤ Omtjob ├──)─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
│ ┌─\ALL────────────┐ │
│ │ ┌──
───────────┐ │ │
(6)
├─SLTUSRID(─── ─┴──6─user-name─┴──
(7)
─┴──)─┤
│ ┌─\NONE───────────┐ │
│ │ ┌──
───────────┐ │ │
(8)
└─OMTUSRID(─── ─┴──6─user-name─┴──
(9)
─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(10)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(11)
─┴──)─┤
│ ┌─\NONE──────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(12)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(13)
─┴──)─┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\ALL──────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(14)
├─SLTSBS(──── ─┴──6─subsystem-name─┴───
(15)
─┴──)─┤
│ ┌─\NONE─────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(16)
└─OMTSBS(──── ─┴──6─subsystem-name─┴───
(17)
─┴──)─┘

A-56 Performance Tools V4R2

 PRTPOLRPT

5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL────────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(18)
├─SLTLINE(──── ─┴──6─communications-line-name─┴───
(19)
─┴──)─┤
│ ┌─\NONE───────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(20)
└─OMTLINE(──── ─┴──6─communications-line-name─┴───
(21)
─┴──)─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\ALL───────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(22)
├─SLTCTL(──── ─┴──6─controller-name─┴───
(23)
─┴──)─┤
│ ┌─\NONE──────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(24)
└─OMTCTL(──── ─┴──6─controller-name─┴───
(25)
─┴──)─┘
5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(26)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(27)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(28)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴───
(29)
─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTPOLRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:
┌─\SELECT────────────────────────────────────────────────────────┐
│ ┌─\FIRST─────┐ ┌─\FIRST─────┐ │
├──┴─┴─start-time─┴──┼────────────┼──┬────────────────────────────┬─┴──)──┤
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─┴─end-time─┴──┼──────────┼─┘
└─end-date─┘
Sltjob:
┌─\ALL────────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(3)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Omtjob:
┌─\NONE───────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(5)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
2 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

3 A maximum of 50 repetitions

4 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

Appendix A. Performance Tools CL Commands A-57

PRTPOLRPT

5 A maximum of 50 repetitions
6 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-
sive.
7 A maximum of 50 repetitions
8 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-
sive.
9 A maximum of 50 repetitions
10 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-
sive.
11 A maximum of 16 repetitions
12 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-
sive.
13 A maximum of 16 repetitions
14 The SLTSBS parameter and the OMTSBS parameter are mutually exclusive.
15 A maximum of 50 repetitions
16 The SLTSBS parameter and the OMTSBS parameter are mutually exclusive.
17 A maximum of 50 repetitions
18 The SLTLINE parameter and the OMTLINE parameter are mutually exclusive.
19 A maximum of 50 repetitions
20 The SLTLINE parameter and the OMTLINE parameter are mutually exclusive.
21 A maximum of 50 repetitions
22 The SLTCTL parameter and the OMTCTL parameter are mutually exclusive.
23 A maximum of 50 repetitions
24 The SLTCTL parameter and the OMTCTL parameter are mutually exclusive.
25 A maximum of 50 repetitions
26 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually
exclusive.
27 A maximum of 50 repetitions
28 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually
exclusive.
29 A maximum of 50 repetitions

Purpose
The Print Pool Report (PRTPOLRPT) command produces a pool-oriented report
from the performance data collected by the Start Performance Monitor
(STRPFRMON) command. The report is written to the printer file QPPTITVP. The
two sections of the report are subsystem activity and workload activity of storage
pools. The information is presented according to interval order. Jobs may be selec-
tively included in, or excluded from, the report based on a variety of job details and
interval times.

Examples
Example 1: Printing a Report
PRTPOLRPT MBR(DTAð71588A)

This command submits a batch job to print a report on all jobs in all intervals in the
member DTA071588A of the performance data files in library QPFRDATA. The
report title is taken from the text of that member.

A-58 Performance Tools V4R2

 PRTRSCRPT

Example 2: Selecting Intervals to Include in Report

PRTPOLRPT MBR(DTAð71588A) PERIOD(\SELECT)

This command submits a job to print a report from the same data, but first shows a
display from which the user interactively selects the intervals to include.

Example 3: Specifying Data Collection Time Period

PRTPOLRPT MBR(DTAð71588A) PERIOD((233ð)(ð13ð))

This command submits a job to print a report on the data collected from 11:30 PM
on the first day of collection through 1:30 AM on the last day of collection.
However, if data collection started and ended on the same day, an error message
is printed, because the specified ending date and time is before the specified
starting date and time.

Example 4: Specifying a User ID

PRTPOLRPT MBR(DTAð71588A) SLTUSRID(D46\)
JOBD(\NONE)

This command interactively prints a report for all jobs with a user ID starting with
D46.

Example 5: Specifying a User ID

PRTPOLRPT MBR(DTAð71588A) SLTJOB(D46\/\N)
JOBD(\NONE)

This command performs the same function as the previous example.

PRTRSCRPT (Print Resource Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTRSCRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬───────────────────────┬─────────5
(1)
│ ┌─\MBRTXT────────┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─TITLE(──┼─\BLANK─────────┼──)─┘
└─'report-title'─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTRSCRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:

Appendix A. Performance Tools CL Commands A-59

PRTSYSRPT

(2)
┌─\SELECT─── ─────────────────────────────────────────────────────┐
│ ┌─\FIRST─────┐ ┌─\FIRST─────┐ │
├──┴─┴─start-time─┴──┼────────────┼──┬────────────────────────────┬─┴──)──┤
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─┴─end-time─┴──┼──────────┼─┘
└─end-date─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
2 *SELECT is valid only in the interactive environment.

Purpose
The Print Resource Report (PRTRSCRPT) command produces a device resource
usage report from the performance data collected by the Start Performance Monitor
(STRPFRMON) command. The report is written to the printer file, QPPTITVR, and
shows device resource information by time interval. Resources may be selected for
inclusion in, or exclusion from, the report based on interval times.

Examples
Example 1: Printing a Report
PRTRSCRPT MBR(DTAð71588A)

This command submits a batch job to print a report on all resources in all intervals
in the member DTA071588A of performance data files in library QPFRDATA. The
report title is taken from the text of that member.

Example 2: Selecting Intervals to Include in Report

PRTRSCRPT MBR(DTAð71588A) PERIOD(\SELECT)

This command submits a job to print a report from the same data, but first shows a
screen from which the user interactively select which intervals to include.

Example 3: Specifying Data Collection Time Period

PRTRSCRPT MBR(DTAð71588A) PERIOD((233ð)(ð13ð))

This command submits a job to print a report on the data collected from 11:30 PM
on the first day of collection through 1:30 AM on the last day of collection.

PRTSYSRPT (Print System Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTSYSRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬───────────────────────┬─────────5
(1)
│ ┌─\MBR───────────┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─TITLE(──┴─'report-title'─┴──)─┘

A-60 Performance Tools V4R2

 PRTSYSRPT

5──┬───────────────────────────────────┬──┬──────────────────────────┬────5
(3)
│ ┌─\ALL────────────────┐ │ ├─SLTJOB(─── ─┤ Sltjob ├──)─┤
(5)
│ │ ┌──
───────────────┐ │ │ └─OMTJOB(─── ─┤ Omtjob ├──)─┘
└─TYPE(──┴──6─┬─\WORKLOAD─┬─┴──
(2)
─┴──)─┘
├─\RSC──────┤
├─\RSCEXPN──┤
├─\POOL─────┤
├─\DISK─────┤
└─\CMN──────┘
5──┬──────────────────────────────────────┬───────────────────────────────5
│ ┌─\ALL────────────┐ │
│ │ ┌──
───────────┐ │ │
(7)
├─SLTUSRID(─── ─┴──6─user-name─┴──
(8)
─┴──)──┤
│ ┌─\NONE────────────┐ │
│ │ ┌──
───────────┐ │ │
(9)
└─OMTUSRID(─── ─┴──6─user-name─┴───
(10)
─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(11)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(12)
─┴──)─┤
│ ┌─\NONE──────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(13)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(14)
─┴──)─┘
5──┬──────────────────────────────────────────┬───────────────────────────5
│ ┌─\ALL──────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(15)
├─SLTSBS(──── ─┴──6─subsystem-name─┴───
(16)
─┴──)─┤
│ ┌─\NONE─────────────────┐ │
│ │ ┌──
────────────────┐ │ │
(17)
└─OMTSBS(──── ─┴──6─subsystem-name─┴───
(18)
─┴──)─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL────────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(19)
├─SLTLINE(──── ─┴──6─communications-line-name─┴───
(20)
─┴──)─┤
│ ┌─\NONE───────────────────────────┐ │
│ │ ┌──
──────────────────────────┐ │ │
(21)
└─OMTLINE(──── ─┴──6─communications-line-name─┴───
(22)
─┴──)─┘
5──┬───────────────────────────────────────────┬──────────────────────────5
│ ┌─\ALL───────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(24)
├─SLTCTL(──── ─┴──6─controller-name─┴───
(25)
─┴──)─┤
│ ┌─\NONE──────────────────┐ │
│ │ ┌──
─────────────────┐ │ │
(26)
└─OMTCTL(──── ─┴──6─controller-name─┴───
(27)
─┴──)─┘
5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(28)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(29)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(30)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴───
(31)
─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTSYSRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘

Appendix A. Performance Tools CL Commands A-61

PRTSYSRPT

Period:
┌─\SELECT────────────────────────────────────────────────────────┐
│ ┌─\FIRST─────┐ ┌─\FIRST─────┐ │
├──┴─┴─start-time─┴──┼────────────┼──┬────────────────────────────┬─┴──)──┤
└─start-date─┘ │ ┌─\LAST────┐ ┌─\LAST────┐ │
└─┴─end-time─┴──┼──────────┼─┘
└─end-date─┘
Sltjob:
┌─\ALL────────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(4)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Omtjob:
┌─\NONE───────────────────────────────────────────────┐
│ ┌──
───────────────────────────────────────────────┐ │
├──┴──6───┬─────────────────────────────┬──job-name───┴──
(6)
─┴────────────────┤
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
2 Specific report section with maximum of 6 repetitions.

3 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the

same time.
4 A maximum of 50 repetitions.

5 The SLTJOB parameter and the OMTJOB parameter cannot be specified at the

same time.
6 A maximum of 50 repetitions.

7 The SLTUSRID parameter and the OMTUSRID parameter cannot be specified at

the same time.

8 A maximum of 50 repetitions.

9 The SLTUSRID parameter and the OMTUSRID parameter cannot be specified at

the same time.

10 A maximum of 50 repetitions.

11 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified

at the same time.

12 A maximum of 16 repetitions.

13 The SLTPOOLS parameter and the OMTPOOLS parameter cannot be specified

at the same time.

14 A maximum of 16 repetitions.

15 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the

same time.
16 A maximum of 50 repetitions.

17 The SLTSBS parameter and the OMTSBS parameter cannot be specified at the

same time.
18 A maximum of 50 repetitions.

19 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the

same time.
20 A maximum of 50 repetitions.

A-62 Performance Tools V4R2

 PRTSYSRPT

21 The SLTLINE parameter and the OMTLINE parameter cannot be specified at the
same time.
22 A maximum of 50 repetitions.
23 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
24 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
25 A maximum of 50 repetitions.
26 The SLTCTL parameter and the OMTCTL parameter cannot be specified at the
same time.
27 A maximum of 50 repetitions.
28 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
29 A maximum of 50 repetitions.
30 The SLTFCNARA parameter and the OMTFCNARA parameter cannot be speci-
fied at the same time.
31 A maximum of 50 repetitions.

Purpose
The Print System Report (PRTSYSRPT) command generates and prints a system
operation overview report from the performance data collected by the Start Per-
formance Monitor (STRPFRMON) command. The report is written to the printer file
QPPTSYSR. The system work load, resource utilization expansion, storage pool
utilization, disk utilization, and communications summary are presented in the
report.

Examples
Example 1: Printing a Report
PRTSYSRPT MBR(APRIL18)

or
PRTSYSRPT MBR(APRIL18) SECTION(\ALL)

These commands print a complete system report for the performance data member
APRIL18 in library QPFRDATA. The report title is the same as the text in the
member.

Example 2: Selecting Intervals to Include in Report

PRTSYSRPT MBR(NOV1) PERIOD(\SELECT)
TITLE('Intervals with Highest Response Times')

This command prints a system report for the data member NOV1 in library
QPFRDATA. The user is presented with the interval-selection screen, which allows
sorting of the intervals according to various criteria and the selection of certain
intervals to be included in the report. The title of the report is "Intervals with Highest
Response Times."

Example 3: Selecting Sections to Include in Report

PRTSYSRPT MBR(NOV1) SECTION(\DSKUTL)

Appendix A. Performance Tools CL Commands A-63

PRTTNSRPT

This command prints only the Disk Utilization section of the system report for the
data member NOV1.

PRTTNSRPT (Print Transaction Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTTNSRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
5──┬───────────────────────────────┬──────────────────────────────────────5
│ ┌─\BLANK─────────┐ │
└─TITLE(──┴─'report-title'─┴──)─┘
(P)
5──┬─────────────────────────────────┬── ──────────────────────────────────5
│ ┌──
──────────────┐ │
│ │ ┌─\SUMMARY─┐ │ │
└─RPTTYPE(───6─┼─\TNSACT──┼─┴──
(1)
──)─┘
├─\TRSIT───┤
├─\FILE────┤
└─\TRCDTA──┘
5──┬────────────────────────────────────────────┬─────────────────────────5
│ ┌─\FIRST─────┐ ┌─\LAST────┐ │
(2)
└─PERIOD(─── ─┴─start-time─┴──┼──────────┼──)─┘
└─end-time─┘
5──┬───────────────────────────┬──┬──────────────────────────────────┬────5
│ ┌──
─────────┐ │ │ ┌─\ALL───────────┐ │
│ │ ┌─\SS─┐ │ │ │ │ ┌──
──────────┐ │ │
└─OPTION(───6─┼─\SI─┼─┴──
(3) (4)
──)─┘ ├─SLTJOB(─── ─┴──6─job-name─┴──
(5)
─┴──)─┤
├─\OZ─┤ │ ┌─\NONE──────────┐ │
├─\EV─┤ │ │ ┌──
──────────┐ │ │
├─\HV─┤ (6)
└─OMTJOB(─── ─┴──6─job-name─┴──
(7)
─┴──)─┘
├─\DI─┤
└─\DQ─┘
5──┬────────────────────────────────────────────────────┬─────────────────5
│ ┌─\ALL─────────────────────────┐ │
│ │ ┌──
────────────────────────┐ │ │
(8)
├─SLTUSRID(─── ─┴──6─┬─user-name──────────┬─┴──
(9)
─┴──)───┤
│ └─generic\-user-name─┘ │
│ ┌─\NONE─────────────────────────┐ │
│ │ ┌──
────────────────────────┐ │ │
(10)
└─OMTUSRID(──── ─┴──6─┬─user-name──────────┬─┴───
(11)
─┴──)─┘
└─generic\-user-name─┘
5──┬─────────────────────────────────────────────────────┬────────────────5
│ ┌─\ALL───────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(12)
├─SLTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(13)
─┴──)─┤
│ ┌─\NONE──────────────────────────┐ │
│ │ ┌──
─────────────────────────┐ │ │
(14)
└─OMTPOOLS(──── ─┴──6─storage-pool-identifier─┴───
(15)
─┴──)─┘
5──┬───────────────────────────────────────────────────┬──────────────────5
│ ┌─\ALL────────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(16)
├─SLTFCNARA(──── ─┴──6─functional-area-name─┴───
(17)
─┴──)─┤
│ ┌─\NONE───────────────────────┐ │
│ │ ┌──
──────────────────────┐ │ │
(18)
└─OMTFCNARA(──── ─┴──6─functional-area-name─┴───
(19)
─┴──)─┘

A-64 Performance Tools V4R2

 PRTTNSRPT

5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTTNSRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬────────────────────────────────────────────────┬────────────────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD──┐ │
└─JOBD(──┬─┼───────────────┼──┴─jobd-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE────────────────────────────┘
Notes:
1 A maximum of 4 report types
P All parameters preceding this point can be specified in positional form.

2 *PERIOD contains two lists of one element each. *N must be specified for any

element that precedes the value(s) to be specified to maintain its position in the
parameter value sequence.
3 A maximum of 5 repetitions

4 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

5 A maximum of 50 repetitions

6 The SLTJOB parameter and the OMTJOB parameter are mutually exclusive.

7 A maximum of 50 repetitions

8 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-

sive.
9 A maximum of 50 repetitions

10 The SLTUSRID parameter and the OMTUSRID parameter are mutually exclu-

sive.
11 A maximum of 50 repetitions

12 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-

sive.
13 A maximum of 16 repetitions

14 The SLTPOOLS parameter and the OMTPOOLS parameter are mutually exclu-

sive.
15 A maximum of 16 repetitions

16 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually

exclusive.
17 A maximum of 50 repetitions

18 The SLTFNCARA parameter and the OMTFNCARA parameter are mutually

exclusive.
19 A maximum of 50 repetitions

Purpose
The Print Transaction Report (PRTTNSRPT) command is used to create and print
performance reports that show detailed information about the transactions that
occurred during the time that the performance data was collected. These reports
use trace data collected by specifying TRACE(*ALL) on the Start Performance
Monitor (STRPFRMON) command. Jobs may be selectively included in the reports
or excluded from the reports based on a variety of job details and interval times.

Appendix A. Performance Tools CL Commands A-65

PRTTRCRPT

Examples
Example 1: Printing a Summary Transaction Report
PRTTNSRPT MBR(TUESAM)

This command produces a summary transaction report. The data input to the report
is all the data that exists in member TUESAM in library QPFRDATA. The request is
sent to batch. The report output is directed to the output queue specified in the job
description, QPFRJOBD.

Example 2: Printing a Transaction Detail Report

PRTTNSRPT MBR(TUESAM) RPTTYPE(\TNSACT)
SLTJOB(WSð1)

This command produces a transaction detail report for the selected job, WS01. The
request is sent to batch. The report output is directed to the output queue specified
in the job description, QPFRJOBD.

PRTTRCRPT (Print Trace Report) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

55──PRTTRCRPT──MBR(──member-name──)──┬───────────────────────────┬────────5
│ ┌─QPFRDATA─────┐ │
└─LIB(──┴─library-name─┴──)─┘
(P)
5──┬───────────────────────────────┬── ──┬───────────────────────┬─────────5
(1)
│ ┌─\MBR───────────┐ │ └─PERIOD(─── ─┤ Period ├─┘
└─TITLE(──┴─'report-title'─┴──)─┘
5──┬────────────────────────┬─────────────────────────────────────────────5
│ ┌─PRTTRCRPT─┐ │
└─JOB(──┼─\MBR──────┼──)─┘
└─job-name──┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Period:
┌─\AVAIL─────┐ ┌─\CURRENT───┐
├──┴─start-time─┴──┼─\BEGIN─────┼──)──┬──────────────────────────────────┬─┤
─
└─start-date─┘ │ ┌─\AVAIL───┐ ┌─\CURRENT─┐ │
└─(──┴─end-time─┴──┼──────────┼──)─┘
├─\END─────┤
└─end-date─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 *PERIOD contains two lists of two elements each. *N must be specified for any

element that precedes the value(s) to be specified, to maintain its position in the
parameter value sequence.

A-66 Performance Tools V4R2

 RMVPEXDFN

Purpose
The Print Trace Report (PRTTRCRPT) command produces a report showing
resources utilized, exceptions, and state transitions for batch jobs traced through
time based on the trace data collected by the Performance Monitor when job type
tracing is requested. This report runs against the specified member (in the
QTRJOBT file of the QPFRDATA library) that was created when the Print Trans-
action Report (PRTTNSRPT) command was run with the *FILE option.

Examples
Example 1: Printing a Job Trace Summary Report
PRTTRCRPT MBR(JUNEð1)

This command submits a batch job that generates a Job Trace Summary report
using the performance data found in the member JUNE01 of file QTRJOBT located
in the default library QPFRDATA. The report covers the entire collection period, and
the title of the report is set to the name of the database file member.

Example 2: Specifying a Report Time Period

PRTTRCRPT MBR(NOV15)
PERIOD(('ð8ðð:ðð' '11/15/89')
('2359:59' '11/15/89'))
TITLE('Job Trace Reports for November 15')

This command submits a batch job that generates a Job Trace Summary report.
The performance data comes from member NOV15 of file QTRJOBT of the default
library QPFRDATA. The report covers the time period 8:00 in the morning to mid-
night of one day.
Note: The format for the date and time is determined by the system values
QDATFMT and, because separators are used in this example, QDATSEP.

RMVPEXDFN (Remove Performance Explorer Definition) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──RMVPEXDFN──DFN(──┬─definition-name──────────┬──)─── ──────────────────5%
├─generic\-definition-name─┤
└─\ALL─────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Remove Performance Explorer Definition (RMVPEXDFN) command removes
one or more performance explorer definitions from the system. Each definition is
stored as a member in the QAPEXDFN file in library QUSRSYS. The member
name is the same as the definition name.

Restriction: The user must have object existence authority for file QAPEXDFN in
library QUSRSYS.

Appendix A. Performance Tools CL Commands A-67

STRJOBTRC

Examples
Example 1: Removing a Single Definition
RMVPEXDFN DFN(SAMPLE)

This command removes the member named SAMPLE from file QAPEXDFN in
library QUSRSYS that contains the performance explorer definition named
SAMPLE.

Example 2: Removing All Definitions that Start with SAM

RMVPEXDFN DFN(SAM\)

This command removes all definitions with names that start with SAM by removing
all members that start with SAM from file QAPEXDFN in library QUSRSYS.

STRJOBTRC (Start Job Trace) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

(P)
55──STRJOBTRC──┬─────────────────────────────────┬── ──────────────────────5
│ ┌─1ð24────────────┐ │
└─MAXSTG(──┴─maximum-K-bytes─┴──)─┘
5──┬────────────────────────────────────────────────────────┬────────────5%
└─JOB(──┬─\─────────────────────────────────────────┬──)─┘
└─┬─────────────────────────────┬──job-name─┘
└─┬─────────────┬──user-name/─┘
└─job-number/─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Start Job Trace (STRJOBTRC) command starts the OS/400 system job tracing
function to collect performance statistics about the specified job. It issues the fol-
lowing command:
TRCJOB MAXSTG(maxstg) TRCFULL(\STOPTRC)

The MAXSTG value used is either the STRJOBTRC default or a specified value.
The STRJOBTRC command issues a Service Job (SRVJOB) command if a job
other than the current job is specified.

Once job tracing is started, a trace record is created for every external (program)
call and return, exception, message, and work station wait in the job. At least two,
and usually more, trace records are created for every I/O statement (open, close,
read, or write) in a high-level language program.

After the target programs have been run, the End Job Trace (ENDJOBTRC)
command is used to turn tracing off, record the collected information in a database
file, and optionally produce reports used to analyze the data. The Print Job Trace
(PRTJOBTRC) command may also be used to print the same report at any time
thereafter. Alternatively, the command

A-68 Performance Tools V4R2

 STRPEX

TRCJOB SET(\CNL)

may be used to cancel the job trace (turn it off without recording any collected
data).

Tracing has a significant impact on the performance of the current job. It also
affects the performance of the system in general, but to a lesser extent.

Example
STRJOBTRC MAXSTG(512)

This command starts collecting trace data for the current job. It uses a 512KB trace
buffer, which is large enough to hold about 7000 trace records.

STRPEX (Start Performance Explorer) Command

Format
Job: B,I Pgm: B,I REXX: B,I Exec

┌─\NEW─────┐
(P)
55──STRPEX──SSNID(──name──)──OPTION──┼─\INZONLY─┼──)─── ───────────────────5
└─\RESUME──┘
5──┬─────────────────────────────┬───────────────────────────────────────5%
(1, 2)
└─DFN(──── ─definition-name──)─┘
Notes:
P All parameters preceding this point can be specified in positional form.
1 DFN parameter must be specified for OPTION(*NEW) and OPTION(*INZONLY).

2 DFN parameter is not valid when OPTION(*RESUME) specified.

Purpose
The Start Performance Explorer (STRPEX) command starts a new performance
explorer session or resumes a suspended performance explorer session.
Note: You are allowed to have only one performance explorer session at a time.
Multiple sessions are not allowed.

Examples
Example 1: Start a New Session
STRPEX SSNID(TESTRUN2)
DFN(NEWDESC) OPTION(\NEW)

This command starts a new session of the performance explorer using the criteria
identified in a definition named NEWDESC. The new session name is TESTRUN2.

Example 2: Resume a Suspended Session

STRPEX SSNID(TESTRUN1)
OPTION(\RESUME)

This is a resume operation for an already existing session of the performance

explorer named TESTRUN1.

Appendix A. Performance Tools CL Commands A-69

STRPFRT

STRPFRG (Start Performance Graphics) Command

Format
Job: I Pgm: I REXX: I Exec

55──STRPFRG──┬──────────────────────────────┬─────────────────────────────5
│ ┌─QPFRDATA─────┐ │
└─FMTLIB(──┴─library-name─┴──)─┘
(P)
5──┬──────────────────────────────┬── ──┬───────────────────────┬──────────5
│ ┌─QPFRDATA─────┐ │ │ ┌─\CMD─────┐ │
└─PFRLIB(──┴─library-name─┴──)─┘ └─JOB(──┼─\MBR─────┼──)─┘
└─job-name─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Start Performance Graphics (STRPFRG) command calls up the performance
graphics menu giving the user access to all available performance graphics. This
command is valid only in an interactive environment.

Example
STRPFRG

This command displays the menu interface for Performance Tools graphics.
Library QPFRDATA is used for storing and retrieving graph formats, graph pack-
ages, and performance data for this session. Any requests that are submitted to run
in batch use job description QPFRJOBD found in the library list.

STRPFRT (Start Performance Tools) Command

Format
Job: I Pgm: I REXX: I Exec

A-70 Performance Tools V4R2

 WRKFCNARA

(P)
55──STRPFRT──┬───────────────────────────┬── ──┬───────────────────────┬───5
│ ┌─QPFRDATA─────┐ │ │ ┌─\CMD─────┐ │
└─LIB(──┴─library-name─┴──)─┘ └─JOB(──┼─\MBR─────┼──)─┘
└─job-name─┘
5──┬───────────────────────────────────────────────────────────┬─────────5%
│ ┌─\LIBL/────────┐ ┌─QPFRJOBD─────────────┐ │
└─JOBD(──┬─┼───────────────┼──┴─job-description-name─┴─┬──)─┘
│ ├─\CURLIB/──────┤ │
│ └─library-name/─┘ │
└─\NONE───────────────────────────────────────┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Start Performance Tools (STRPFRT) command calls the Performance
Tools/400 menu interface. This shows the Performance Tools main menu
(PERFORM), which provides access to all of the Performance Tools.

Example
STRPFRT

This command calls the menu interface to Performance Tools. The library,
QPFRDATA, is used in storing and getting the performance data for this session.
Requests that are submitted to run in batch use the job description QPFRJOBD in
the library list.

WRKFCNARA (Work with Functional Areas) Command

Format
Job: I Pgm: I REXX: I Exec

(P)
55──WRKFCNARA──┬───────────────────────────┬── ───────────────────────────5%
│ ┌─QPFRDATA─────┐ │
└─LIB(──┼──────────────┼──)─┘
└─library-name─┘
Note:
P All parameters preceding this point can be specified in positional form.

Purpose
The Work with Functional Areas (WRKFCNARA) command allows the user to
create, change, copy, and delete functional areas at the menu level.

Example
WRKFCNARA

This command allows the user to create, change, and delete functional areas at the
menu level. The library, QPFRDATA, is used for storing and retrieving functional
areas for this session.

Appendix A. Performance Tools CL Commands A-71

WRKSYSACT

WRKSYSACT (Work with System Activity) Command

Format
Job: I Pgm: I Exec

(P)
55──WRKSYSACT──┬───────────────────────┬── ────────────────────────────────5
│ ┌─\─────┐ │
└─OUTPUT(──┼─\FILE─┼──)─┘
└─\BOTH─┘
5──┬───────────────────────────────────┬──┬───────────────────┬───────────5
│ ┌─5───────────────┐ │ │ ┌─\CPU─┐ │
└─INTERVAL(──┴─interval-length─┴──)─┘ └─SEQ(──┴─\IO──┴──)─┘
5──┬─────────────────────────┬────────────────────────────────────────────5
│ ┌─\ALL───┐ │
└─INFTYPE(──┼─\JOBS──┼──)─┘
└─\TASKS─┘
5──┬─────────────────────────────────────┬────────────────────────────────5
│ ┌─6ð──────────────────┐ │
└─NBRITV(──┴─number-of-intervals─┴──)─┘
5──┬──────────────────────────┬──┬───────────────────────────┬───────────5%
│ ┌─QAITMON─────┐ │ │ ┌─QPFRDATA─────┐ │
└─MBR(──┴─member-name─┴──)─┘ └─LIB(──┴─library-name─┴──)─┘
Note:
PAll parameters preceding this point can be specified in positional form.

Purpose
The Work with System Activity (WRKSYSACT) command allows you to work inter-
actively with the jobs and tasks currently running in the system. Besides having the
capability to view this data on the display station, the user may also store the data
in a database file.

Restriction:

The user must have *JOBCTL authority to use this command.

Examples
Example 1: Working with Jobs and Tasks
WRKSYSACT

This command displays both jobs and tasks in descending order of processing time
used.

Example 2: Working with Jobs Only

WRKSYSACT OUTPUT(\BOTH) INFTYPE(\JOBS)
MBR(JOBACT)

This command displays jobs only and writes this information to member JOBACT of
database file QAITMON located in default library QPFRDATA.

A-72 Performance Tools V4R2

Appendix B. Defining Transaction Boundaries
Performance tools reports show different values for transaction service time and
resource use, depending on what command you use to analyze the performance
data. These values vary because of differences in the data collected by the com-
mands, and can supply different values for the transaction boundary start and end
times. Be careful when you analyze and compare data collected for the same run
using different tools.

Some of the commands used in this appendix are available only in the Manager
feature. Appendix D, Comparison of Performance Tools, provides additional infor-
mation about Performance Tools functions.

Elements of Response Time

The elements of end-user (external) response time to interactive transactions are
composed of communications time (input and output) and host (internal) response
time, as shown in Figure B-1. For locally attached display stations, communications
time managers the local Work Station Input/Output Processor (IOP) time. For
remotely attached display stations, communications time includes communications
line time, communications IOP time, and Remote Work Station Controller time as
appropriate.

%──────────────── End-User Response Time ─────────────────5

(External Response Time)
│ │ │ │
│─────────────┼────────────────────────────┼──────────────┤
│ │ │ │
Input Host Response Time Output
Communications (Internal Response Time) Communications
Time Time

Figure B-1. Elements of Interactive Response Time

The AS/400 system interactive response time values reported by the Work with
Active Job (WRKACTJOB), Print System Report (PRTSYSRPT), Print Component
Report (PRTCPTRPT), and Print Transaction Report (PRTTNSRPT) commands
refer only to the host (internal) response time. (An exception to this is the Local
Work Station Report, shown in Figure 7-20 on page 7-33. This report does factor
in local work station IOP time.)

For locally attached displays, the communications time is usually less than 1
second. For remote displays, the communications time may be longer. To approxi-
mate the actual time, use the line speed and number of characters sent and
received, assuming that the line is not heavily loaded. If the line is heavily loaded,
the external response time increases due to the queuing time. Review the line
utilization and data transmission values on the System Report, shown in “What Is
the System Report?” on page 7-17 to determine line component to approximate
line time.

 Copyright IBM Corp. 1998 B-1

 You can also use the BEST/1 capacity planning tool to estimate external response
times at local and remote display stations, because BEST/1 projects both internal
and external response times. BEST/1 supports only 5250-type devices that are
attached either locally through twinaxial cable, remotely through SDLC communi-
cations lines, or through LAN-attached work stations.

The host response time can, however, be shown in more detail, as in Figure B-2.

%─────────────────── Host Response Time ─────────────────5

(RT)
│ │ │ │
│ │ │ │
├───────────┼─────────────────────────────────────┼───────────┤
│ │ │ │
│ │ │ │
Ineligible │ Active Time │ EX-Wait
Time │ │ Time
(IT) │ │ (EW)
│ │ │ │ │ │
└─────┼───────┼──────┼───────┼────────┘
│ │ │ │
CPU Wait Disk Wait Wait
Time for CPU I/O for I/O in MPL
(CT) (CW) (DT) (DW) (MW)

Figure B-2. Elements of Host Response Time

Note: Multiprogramming Level (MPL) is a term used interchangeably with Activity

Level.

The average ineligible time, processing unit time, wait in MPL time, and exceptional
wait time per transaction are available directly from the output of the PRTTNSRPT
command.

Differences in the Transaction Response Reports

Figure B-3 on page B-3 compares the ways that the Print Job Trace
(PRTJOBTRC) command, the PRTTNSRPT command, the PRTSYSRPT
command, and the WRKACTJOB command determine transaction boundaries.

B-2 Performance Tools V4R2

 PRTTNSRPT
PRTJOBTRC
( W- > A )
or
(W->I) (I->A) (A->W)

1 2 3 4 5 6 7 8 9 10
(WRITE*)

PRTSYSRPT and WRKACTJOB

Work Station I/O Manager

(1) External I/O request received (PRTSYSRPT start)
(2) Licensed internal code processing complete
(3) Job put into activity level or ineligible state
(4) Trace record generated (PRTTNSRPT start)

OS/400 System Application

(5) Ineligible time complete (I-A)
(6) Return to QWSGET (Start of transaction
on job trace)
(7) Write to Work Station

Work Station I/O Manager

(8) Call QT3REQIO (End of transaction on job trace,
Transaction response times, PRTSYSRPT
transaction end)
(9) Job goes to IOM to wait on I/O (PRTSYSRPT
transaction end)
(10) A-W trace recorded (PRTTNSRPT transaction end)
RV2S065-2

Figure B-3. Comparison of Transaction Boundary Definitions

PRTSYSRPT and WRKACTJOB define a transaction from the time it is processed

by the Licensed Internal Code I/O manager (Licensed Internal Code IOM) until the
system work station I/O program QT3REQIO is called to wait for input.

PRTTNSRPT defines a transaction from the time trace records are produced at the
beginning when the job state changes from wait-to-active or wait-to-ineligible (the
start) until the job goes to a long wait (active-to-wait).
Note: Values *DI and *DQ on the OPTION parameter use existing transaction
boundary trace records to count transactions instead of the wait-to-active
transition.

These commands include the time the job spent in the ineligible state waiting for an
activity level as part of the transaction response time.

PRTJOBTRC defines a transaction from the time the job becomes eligible (for
example, it is granted an activity level) within the system work station input program
(QWSGET), until the system work station I/O program QT3REQIO is called to wait
for input.
Note: This command does not include the time spent in the ineligible state waiting
for an activity level in the transaction boundary definition.

Appendix B. Defining Transaction Boundaries B-3

Operational Considerations
Limitations exist in the system’s ability to detect certain types of transactions.

When you review performance reports, be aware of when your system workload
consists of any of the following types of work:
Ÿ Programmable work station servers
Ÿ Distributed data management (DDM) servers
Ÿ 3270 emulation jobs
Ÿ Finance terminals
Ÿ Pass-through jobs

Transaction-type data (such as the data collected for throughput and response
time) is unavailable sometimes, and in some instances (such as for finance types of
work), cannot be associated with the individual jobs or terminals that originated the
transactions.

When you find that differences exist between the sample data reports
(PRTSYSRPT or PRTCPTRPT) and the trace data report (PRTTNSRPT), it is often
due to the presence of one or more of these types of work. Use the Select/omit
option on the reporting commands to remove these types of jobs so the information
shown on the reports is more representative of your environment.

You may find that the performance tools transaction information is inaccurate for
applications such as RM/COBOL-85 for the AS/400 licensed program that do field-
by-field processing. (Field-by-field processing implies that for every field in which
data is entered, there is processing by the CPU as the field is exited.) The tools
report each field processed as a transaction. Because these ‘field’ transactions may
not do much processing other than return to the screen to enable the next field to
be entered, the transaction information is skewed. When all of the fields on the
screen have been entered, what would be viewed as a normal transaction occurs,
that is, all of the information is processed.

If the transaction information is skewed due to field by field processing, it cannot be

used as input to BEST/1. BEST/1 uses the transaction information to establish its
base information. It then uses the base information to predict the AS/400 modes,
response time, transactions, and utilizations for a given number of work stations. If
the transaction information is skewed, BEST/1 may give incorrect results.

B-4 Performance Tools V4R2

Appendix C. Correlation of the System/36 and AS/400 System
Performance Parameters—Manager Feature
The differences that exist between the AS/400 system and System/36, both hard-
ware and software, make it difficult to map performance parameters from one
system to the other. Table C-1 shows the correlation for some of the key
System/36 performance parameters. This information is intended for those new
AS/400 system users who have a background in System/36 performance analysis.

Table C-1 (Page 1 of 2). Correlation of the System/36 and AS/400 System Performance Parameters
System/36 Parameter AS/400 Parameter
Main storage processor These parameters correlate to the processor utilization reported by Perform-
Control storage processor ance Tools.
See “Resource Utilization” on page 7-19 to see the format of this report. This
is a portion of the System Report that is produced when you use the Print
System Report (PRTSYSRPT) command.
Disk (1-4) The System/36 disk information correlates to the disk utilization.
See “Disk Utilization” on page 7-19 to see the format of this report. This is a
portion of the System Report that is produced when you use the PRTSYSRPT
command.
There is also a Disk Activity Report that is a portion of the Component Report.
You produce the Component Report when you use the Print Component
Report (PRTCPTRPT) command.
Communications lines (1-8) See “Communications Summary” on page 7-19 to see the format of this report.
This is a portion of the System Report that is produced when you use the
PRTSYSRPT command.
More detailed information is shown in the Communications Report, which is a
portion of the Component Report. You produce the Component Report when
you use the PRTCPTRPT command.
Details about communications lines are in a Resource Interval Report: Commu-
nications Line Summary report produced by the Print Resource Report
(PRTRSCRPT) command (see “Resource Interval Report” on page 7-76 for
more information).
Communications lines information can be displayed by using the Display Per-
formance Data (DSPPFRDTA) command.
Translated transfer calls/loads Nothing on the AS/400 system correlates to these counters.
User Area Disk Activity The closest correlation is shown in “Storage Pool Utilization” on page 7-19.
(UADA)
This report is a portion of the System Report that is produced when you use
the PRTSYSRPT command.
The faults per second statistics represent the measure of main storage utiliza-
tion.
More detailed information can be found in “Storage Pool Activity” on
page 7-26. This report is a portion of the Component Report that is produced
when you use the PRTCPTRPT command.

 Copyright IBM Corp. 1998 C-1

 Table C-1 (Page 2 of 2). Correlation of the System/36 and AS/400 System Performance Parameters
System/36 Parameter AS/400 Parameter
Number of disk record waits On the AS/400 system, the corresponding measure is given by seize/lock sta-
tistics.
To obtain seize/lock statistics, you must collect trace data when running the
performance monitor. To do this, specify TRACE(*ALL) on the Start Perform-
ance Monitor (STRPFRMON) command.
This information is shown as a portion of the special summary reports that are
part of the Transaction Report. To produce it, leave the OPTION(*SS) value on
the PRTTNSRPT command.
A more detailed analysis can also be made by using the Print Lock Report
(PRTLCKRPT). See Chapter 10, Performance Utilities—Manager Feature for
information on this command.
Work station controller utiliza- The local work station controller is included in the IOP Utilizations Report.
tion
This report is a portion of the Component Report that is produced when you
use the PRTCPTRPT command. The IOP Utilizations Report includes the con-
troller time in the response times that it shows.
File placement File placement on the AS/400 system has little correlation to that of System/36.
The AS/400 operating system, using algorithms to internally partition files,
automatically spreads individual files across all available disk drives to balance
disk utilization. This removes the need to do any file placement analysis.
Disk seek operations greater Because of the file placement actions taken by the AS/400 operating system,
than one-third of the disk there is essentially no user control that can be applied to this measure (other
than adding auxiliary disk space).
The statistic can be found as part of the Component Report.

C-2 Performance Tools V4R2

Appendix D. Comparison of Performance Tools
This appendix compares the functional capability of the Manager feature and the
Agent feature. It specifically notes the differences in the supported menu options
and performance commands. Appendix E, Managing AS/400 System Performance
in a Network, describes an example of a customer using the Performance Tools
Manager feature at a central site and Performance Tools Agent feature at a remote
site.

Comparison of Functions, Menu Options, and Commands

The Agent feature of Performance Tools provides functions to simplify the col-
lection, management, online display, data reduction, and analysis of performance
data on the AS/400 system. Also included in the Agent feature is the performance
explorer tool (performance utilities). The major Performance Tools functions not
contained in the Agent feature are performance and trace reports, capacity plan-
ning, performance utilities (job traces and select file and access group), system
activity monitoring, and performance graphics.

If you require analysis of trace data, viewing data graphically, viewing system
activity in real time, or managing and tracking system growth, the Manager feature
of the Performance Tools licensed program is more useful.

Table D-1 on page D-2 shows the Performance Tools menu options supported by
the Agent feature.

 Copyright IBM Corp. 1998 D-1

 Table D-1. Comparison of Menu Options
Performance Tools Menu Options Agent Feature
1. Select Type of Status No
2. Collect Performance Data Yes
3. Print Performance Reports No
1. System Report
2. Component Report
3. Transaction Report
4. Lock Report
5. Job Report
6. Pool Report
7. Resource Report
8. Batch Job Trace Report
4. Capacity Planning No
1. Create Best Model
5. Performance Utilities
1. Work with Job Traces
2. Work with Performance Explorer
3. Select File and Access Group
6. Configure & Manage Tools
1. Work with Functional Areas
2. Delete Performance Data
3. Copy Performance Data Yes
4. Convert Performance Data Yes
7. Display Performance Data Yes
8. System Activity Yes
9. Performance Graphics
1. Graph Formats No
2. Work with Historical Data Yes
3. Display Graphs and Packages No
10. Advisor Yes

Table D-2 shows performance-related commands, and indicates whether they are
part of OS/400, part of the Manager feature of Performance Tools, or part of the
Agent feature of Performance Tools.

Table D-2 (Page 1 of 3). Comparison of Performance Commands

Manager Agent
Command Description OS/400 Feature Feature
ADDPEXDFN Add performance explorer definition X
ADDPFRCOL Add performance collection X
ANZACCGRP Analyze variable and file usage X
ANZBESTMDL Analyze BEST/1 model X
ANZDBF Analyze files to be used by a program X
ANZDBFKEY Analyze logical to physical database file X
relationships
ANZPFRDTA Advisor X X
ANZPGM Analyze files used by a program X
CHGFCNARA Change functional area X
CHGGPHFMT Change graph format X
CHGGPHPKG Change graph package X

D-2 Performance Tools V4R2

 Table D-2 (Page 2 of 3). Comparison of Performance Commands
Manager Agent
Command Description OS/400 Feature Feature
CHGPEXDFN Change performance explorer definition X
CHGPFRCOL Change performance collection X
CPYFCNARA Copy functional area X
CPYGPHFMT Copy graph format X
CPYGPHPKG Copy graph package X
CPYPFRDTA Copy performance data X X
CRTBESTMDL Create BEST/1 model X X
CRTFCNARA Create functional area X
CRTGPHFMT Create graph format X
CRTGPHPKG Create graph package X
CRTHSTDTA Create historical data X X
CVTPFRDTA Convert performance data from one X
release to another
| CVTPFRTHD Convert performance data from thread- X
| level data to job-level data
DLTBESTMDL Delete BEST/1 model X X
DLTFCNARA Delete functional area X
DLTGPHFMT Delete graph format X
DLTGPHPKG Delete graph package X
DLTHSTDTA Delete historical data X X
DLTPEXDTA Delete performance explorer data X
DLTPFRDTA Delete performance data X X
DMPTRC Dump trace data X
DSPACCGRP Display variable and file usage for jobs X
DSPHSTGPH Display historical graph X
DSPPFRDTA View STRPFRMON sample data X X
DSPPFRGPH Display performance graph X
ENDJOBTRC End job data collection activity X
ENDPEX End Performance Explorer X
ENDPFRCOL End performance collection X
ENDPFRMON Stop performance monitor X
PRTACTRPT Print activity report X
PRTCPTRPT Print component report X
PRTJOBRPT Print job report X
PRTJOBTRC Print job trace report X
PRTLCKRPT Print lock report X
PRTPEXRPT Print performance explorer report X X
PRTPOLRPT Print pool report X

Appendix D. Comparison of Performance Tools D-3

 Table D-2 (Page 3 of 3). Comparison of Performance Commands
Manager Agent
Command Description OS/400 Feature Feature
PRTRSCRPT Print resource report X
PRTSYSRPT Print system report X
PRTTNSRPT Print transaction report X
PRTTRCRPT Print batch job data collected by trace X
RMVPEXDFN Remove performance explorer defi- X
nition
STRPEX Start performance explorer X
STRBEST Capacity planning model X
STRJOBTRC Start job trace X
STRPFRCOL Start performance collection X
STRPFRG Start performance graphics X
STRPFRMON Start performance monitor X
STRPFRT Start Performance Tools X X
WRKACTJOB Job performance data X
WRKDSKSTS Disk space and busy X
WRKFCNARA Work with function areas X
WRKPFRCOL Work with performance collection X
WRKSYSACT Display or record task CPU and disk X X
usage
WRKSYSSTS Memory demand and workload rate X

D-4 Performance Tools V4R2

Appendix E. Managing AS/400 System Performance in a
Network
The following is an example of fictitious Customer A who is managing the growth
and performance of each of the 21 AS/400 systems in the company's network.
Customer A has a Model D60 at the central site and D10 and D25 models at
numerous remote, worldwide sites. Customer A is interested in managing and mon-
itoring the following:
Ÿ Transaction throughput
Ÿ Processing unit utilization, by job type and priority
Ÿ Communications line utilization and error rates
Ÿ Disk utilization and occupancy for each AS/400 system

To meet these objectives, Customer A plans to use the performance monitor and
automatic data collection, which are both features of the OS/400 operating system.
In addition, they will use the Performance Tools licensed program at the central site
AS/400 system and on remote systems.

Planning the Strategy

Customer A has planned the following strategy:
Ÿ Set up automatic data collections at each site, which will include:
– Performance data being collected
– Historical data being created and transmitted to the central site
– The advisor being run, with the results transmitted to the central site
– Data, which was collected more than a month ago, being deleted
Ÿ Create monthly reports at the central site.
Note: These reports will be created using the performance graphics function
of Performance Tools.

Implementing the Strategy

Customer A will first set up automatic data collection at the central site by using the
Work with Performance Collection (WRKPFRCOL) command to set up data col-
lections for Monday’s accounts receivables processing, Wednesday afternoon’s
“normal” workload, and for Friday’s accounts payable processing. These are the
three key workloads that Customer A wants to monitor. For each of these, Cus-
tomer A specifies the day of the week, time of day to start data collection, and a
unique library, using the defaults that are set for two-hour data collections and
15-minute samples.

Once Customer A sets up the data collections at the central site, the following files
are sent to each remote system:
QPFRDATA/QAPMSCOL
QPFRDATA/QAPMSSCD
QPFRDATA/QAPMLSCD

 Copyright IBM Corp. 1998 E-1

 Customer A then sets up job QGPL/QPFRCOL as an automatic start job on each
system (QGPL/QPFRCOL is the default for QBASE and QCTL).

When Customer A is done setting up the data collections, sending the files, and
setting up the automatic start job, the performance monitor schedule is completed.
Note: If you do not already have a job scheduling function, you may be able to
use OfficeVision calendar, the sample scheduler in QUSRTOOL, or a
simple CL program to provide the function to automate this process.

At the remote site, job scheduling is used to schedule an analysis, data reduction,
and a control language (CL) program to transmit data to the central site (see
Figure E-1 on page E-3 for a sample).

E-2 Performance Tools V4R2

START: PGM PARM(&LIB &TOPFRMGR &TOHOST);
/\ Runs the ANZPFRDTA (Advisor) and CRTHSTDTA (Create \/
/\ Historical Data) for the collected data. It then \/
/\ sends the results to the specified Userid at the \/
/\ specified central site. \/

/\ At the central site rename the historical data

/\ files and add the data to the library for each \/
/\ remote system name using RCVNETF. For example: \/
/\ RCVNETF FROMFILE(QAPGHSTD) \/
/\ TOFILE(systemname/\FROMFILE) + \/
/\ FROMMBR(systemname) TOMBR(QAPGHSTD) \/

DCL VAR(&LIB); TYPE(\CHAR) LEN(1ð)

/\ collection library \/
DCL VAR(&TOPFRMGR); TYPE(\CHAR) LEN(1ð)
/\ Host User ID \/
DCL VAR(&TOHOST); TYPE(\CHAR) LEN(1ð)
/\ central node ID \/
DCL VAR(&MBR); TYPE(\CHAR) LEN(1ð)
/\ most recent collection\/
DCL VAR(&SYSNAME); TYPE(\CHAR) LEN(1ð)
/\ system name \/

MONMSG MSGID(CPFðððð)
/\ Monitor for anything +
customize to local requirements \/

RTVMBRD FILE(&LIB/QAPMSYS); MBR(\LAST)

RTNMBR(&MBR);

RTVNETA SYSNAME(&SYSNAME); /\ Get unique system +

name to differentiate files
when received at central site\/

ANZPFRDTA MBR(&MBR); LIB(&LIB); OUTPUT(\PRINT)

/\ Analyze most recently +
collected data in library \/

SNDNETSPLF FILE(QPAVPRT) TOUSRID((&TOPFRMGR &TOHOST)); +

SPLNBR(\LAST) DTAFMT(\ALLDATA)

/\ Create historical data and summary files. \/

/\ The data is put in a new mbr because it will \/
/\ be accumulated at central site rather than this \/
/\ site. BE SURE to RCV the file with MBROPT \ADD \/
CRTHSTDTA MBR(&MBR); LIB(&LIB); JOBD(\NONE)

RMVM FILE(&LIB/QAPGHSTD); MBR(&SYSNAME);

/\ clean up old data \/
RMVM FILE(&LIB/QAPGHSTI); MBR(&SYSNAME);
/\ clean up old data \/

/\ Rename historical data so that a unique \/

/\ name is recognized back at the central site \/
/\ for later review and analysis \/
RNMM FILE(&LIB/QAPGHSTD); MBR(QAPGHSTD)
NEWMBR(&SYSNAME);

Appendix E. Managing AS/400 System Performance in a Network E-3

 RNMM FILE(&LIB/QAPGHSTI); MBR(QAPGHSTI)
NEWMBR(&SYSNAME);

/* Send the SUMMARIZED Jobs and IOP data */

SNDNETF FILE(&LIB/QAPGHSTD);
TOUSRID((TOPFRMGR &TOHOST));
SNDNETF FILE(&LIB/QAPGHSTI);
TOUSRID((TOPFRMGR &TOHOST));
ENDIT: ENDPGM
Figure E-1. CL Program to Collect and Send Historical Data to the Central Site

Also at the remote site, Customer A uses job scheduling to schedule a CL program
to regularly remove “old” performance data (see Figure E-2 for a sample).

At the central site, the job schedule function schedules the following:
Ÿ A CL program to receive and collect data from remote sites
Ÿ A CL program to print advisor reports from remote sites
Ÿ The regular generation of monthly historical graph packages, for example:
DSPHSTGPH QIBMPKG LIB (remotesystemname)
OUTPUT(\PRINT) TYPE (\GPHPKG)
Ÿ A CL program to perform a regular cleanup of “old” historical data collections

START: PGM PARM(&LIB);

/\ Remove first (oldest) data collection. \/
/\ Done regularly, starting X days after first \/
/\ collection and after each collection thereafter \/
/\ the collection the library will maintain a \/
/\ manageable size \/
DCL VAR (&LIB); TYPE (\CHAR) LEN(1ð)
/\ collection library \/
DCL VAR (&MBR); TYPE (\CHAR) LEN(1ð)
/\ oldest collection \/
MONMSG MSGID (CPFðððð)
/\ Monitor everything–customize as needed\/
RTVMBRD FILE (&LIB/QAPMSYS); MBR(\FIRST)
RTNMBR (&MBR);
DLTPFRDTA MBR (&MBR); LIB(&LIB); JOBD(\NONE)
ENDIT: ENDPGM
Figure E-2. CL Program to Regularly Remove Old Performance Data

Additional Capabilities
After reviewing the advisor output and the historical graphs at the central site, Cus-
tomer A may decide that additional detail is needed. This can be done several
ways:
Ÿ To get more detailed graphics, Customer A could send the appropriate
QAPMCONF and QAPGSUMD file members to the central site for processing.
Using this method, Customer A could produce performance graphics of types
*JOBTYPE, *PRIORITY, or *ALL. However, this method does not allow Cus-
tomer A the capability of using select/omit criteria or functional areas.

E-4 Performance Tools V4R2

Ÿ If Customer A wants more detailed reports, trace analyses, full modeling, or full
graphical functions, the remote systems can create a library (using the Copy
Performance Data (CPYPFRDTA) command) containing only the data from that
particular day. This library can then be transmitted to the central system or
saved to tape and sent to the central system. The central system can use the
Performance Tools program to process the data as if it were collected locally.

Appendix E. Managing AS/400 System Performance in a Network E-5

E-6 Performance Tools V4R2
Appendix F. Performance Checklist—Manager Feature
You may find these checklists useful for planning system performance.

Planning for Performance and Tuning

Ÿ Provide proper training:
– OS/400 structure, tailoring, basic tuning
– AS/400 performance analysis and capacity planning
Ÿ Set measurement criteria:
– Define performance objectives
– Set goals
– Take measurements (will you measure peaks or averages?)
– Review measurements
Ÿ Analyze the data.
This requires an understanding of:
– AS/400 commands for collecting data
– Performance Tools programs and reports
– Parameters that affect performance on the AS/400 system, such as:
- Storage pool size, paging
- Size of the process access group (PAG) and purge
- Activity levels
- Time slice
- Job states and transitions
Ÿ Schedule performance review meetings—as often as required to review log
entries and trends.
Ÿ Tune the system using the QPFRADJ (performance adjust) system value. The
values could be 0, 1, 2, or 3.
0 QPFRADJ is off
1 QPFRADJ adjusts shared pools at IPL only
2 QPFRADJ automatically tunes the system at IPL and continually
3 QPFRADJ tunes continually, but not at IPL
QPFRADJ compares system performance to the IBM guidelines every 20
seconds. If it is not within the guidelines on three consecutive comparisons,
QPFRADJ changes the pool sizes, activity level, or both.
Note: You may want to complete QPFRADJ tuning (and then set to 0) before
you run Performance Tools.
Ÿ Print the error log (PRTERRLOG) for hardware problems, and start the system
service tools (STRSST command) to display errors.

 Copyright IBM Corp. 1998 F-1

 Basic Tuning
You can choose to let the system tune itself dynamically (QPFRADJ system value
set to 2 or 3), or you can tune it manually. To tune manually:
Ÿ For initial tuning before you begin performance analysis, compare the pool size
and activity levels to the performance guidelines.
Ÿ After you complete the initial system tuning:
– Evaluate all changes by measuring.
– Make one change at a time.
Ÿ Use the OS/400 CL commands:
– WRKJOB (Work with Job)
– WRKSYSSTS (Work with System Status)
– WRKACTJOB (Work with Active Job)
– WRKDSKSTS (Work with Disk Status)
Ÿ If you have Performance Tools installed, use the WRKSYSACT (Work with
System Activity) command.
Note: This command requires Performance Tools. It is an efficient way to
display currently active jobs and Licensed Internal Code tasks that used
CPU or disk I/O operations since the last time the display was
refreshed.
– It can monitor an individual job.
– One job on the system can use the command.
Ÿ Start the performance monitor to collect data. (Use the Start Performance
Monitor (STRPFRMON) command.)
Note: The performance monitor runs without the Performance Tools program.
However, Performance Tools is needed to create the reports.
– To collect sample data, suggested parameters on this command are:
- Specific member name
- 5-minute interval
- Trace data *NO
- Dump *YES
The duration depends on the requirements, but in general, the performance
monitor should run long enough to collect information that is meaningful.
– Trace data is collected for detailed performance problem analysis.

| Work with System Status Tips

| The Work with System Status identifies page faulting and wait-to-ineligible transi-
| tions for each main storage pool.
| Ÿ For interactive pools, typically you want the wait-to-ineligible transition values to
| be very small (less than 10% of the active-to-wait value). If you see any wait-to-
| ineligible value at all, increase the MAXACT value by 5 to 10 until the wait-to-
| ineligible is 0. Remember to press F10 to reset the statistics. Wait 10 seconds
| between refreshes.

F-2 Performance Tools V4R2

| Ÿ The machine pool (pool 1) should have fewer than 10 faults per second, which
| is the sum of the database and nondatabase faults. You can ignore the Pages
| column.
| Ÿ If only system jobs and subsystem monitors are running in *BASE, then the
| fault rate for that pool should be less than 30 faults per second.
| Ÿ The basic method for tuning your storage pools is to move storage from pools
| with good performance to pools with bad performance. In this situation, you
| should measure performance in response time or as throughput. Continue to
| move storage until the pool with the bad performance gets better, or until the
| pool with the good performance gets worse. Do not decrease a pool by more
| than 10% at a time.
| Ÿ Determining a good fault rate for the user pools can be complicated and will
| vary from pool to pool and from system to system. The fault rate alone is not
| necessarily a measure of good performance or bad performance.
| – For interactive jobs, look at how much the faults contribute to the end
| response time (faults/transaction \ disk response).
| – For batch jobs, look at how much the faults add to the elapsed time of the
| job (total faults \ disk response).
| Ÿ Use the following formula to calculate the approximate number of page faults
| per transaction:
| [(db faults + ndb faults) / active->wait] X 6ð

| Work with Disk Status Tips

| The Work with Disk Status shows the percentage of space used and the per-
| centage of time that the disk arms are busy. If the average percent for the Busy
| column is over 50%, you may need to install more disk arms. The suggestion to
| install more disk arms assumes that any page faulting problems have been
| addressed.

| Work with System Activity Tips

| The Work with System Activity display lists active jobs that have used any CPU in
| the last few seconds. The list is sorted in the order of the amount of CPU seconds
| that are used, which is the default view. If a high priority job (low number) is using
| a lot of CPU (>50%) for an extended period of time, then that job could cause the
| entire system to have poor response times. Here are some suggestions for
| improving your response time in this situation:
| Ÿ If a job or a small set of jobs seems to be using a large percentage of the
| CPU, check the job priority (PTY). If the priority of the job is a lower number
| than the jobs with poor performance, you may want to consider changing the
| priority of the offending job or jobs. Use option 5 (Work with job), then option
| 40 (Change job), and specify a larger RUNPTY value (greater than the priority
| of the jobs that you want to ran faster).
| Ÿ If the offending job is an interactive job that is running a job that is better suited
| to run in batch mode, you may want to contact the user and recommend one of
| the following:
| – That they submit their work as a batch job
| – That they change the priority of the job to 50. 50 is the typical priority for a
| batch job.

Appendix F. Performance Checklist—Manager Feature F-3

| Ÿ If the CPU utilization is high (>80%) and all jobs seem to have an equal, but
| small, CPU percent, this situation could mean that you have too many active
| jobs on the system.

General Tuning Tips

Ÿ Favor output over input.
– The activity level should not be too high.
– Allow the work to finish.
Ÿ Do not mix different types of jobs and priorities in the same pool. (For example,
do not mix class entries in subsystem descriptions.)
Ÿ Remove batch jobs from *BASE by creating another batch pool.
– Route batch jobs to *SHRPOOL1. One batch job to a pool is ideal. (Job
description (JOBD) for routing data; subsystem description (SBSD) for
routing entry).
Ÿ Remove programmers from the interactive pool (*INTERACT) by creating
another interactive pool, *SHRPOOL2 (job description (JOBD) for routing data,
subsystem description (SBSD) for routing entry).

General Performance Facts

Ÿ The size of the process access group (PAG) affects memory utilization. Use the
Display Access Group (DSPACCGRP) command to see the current size.
Ÿ You can collect two types of data in Performance Tools:
– Sample Data allows you to print the following reports:
- Advisor Report
- System Report (Workload; Resource Utilization; Resource Utilization
Expansion; Storage Pool Utilization; Disk Utilization; Communication
Summary)
- Component Report (Interval Activity; Job Workload; Storage Pool; Disk
Activity; IOP Utilization; Local Workstation/Resp time)
– Trace Data allows you to print the following reports:
- Advisor Report
- Summary, Transaction, Transition Reports (more detail about trans-
actions)
Ÿ QTOTJOB system value (QADLTOTJ amount added after original amount).
– Allocates space at IPL
– Sets permanent job structures (work control block table)
Note: QTOTJOB system value should be set at 10% higher than the highest
number of total jobs in the system. Use the Work with System Status
(WRKSYSSTS) command to display the jobs. Leave the value for
QADLTOTJ as it is.
Ÿ QACTJOB system value (QADLACTJ amount added after original amount).
– Space for temporary job structures allocated at IPL.

F-4 Performance Tools V4R2

 – Should be set at highest active job number found. Use the Work with Active
Jobs (WRKACTJOB) command to display. Leave the value for QADLACTJ
as it is.
Ÿ QJOBSPLA remains unchanged.
Ÿ If you are in a Client Access environment with the QPFRADJ system value set
to 2, the machine pool could be adjusted too low.
Ÿ Logical database I/O is one indicator of job activity. Batch or job run time
depends on the CPU time, number of disk operations, and number of excep-
tional waits.

Appendix F. Performance Checklist—Manager Feature F-5

F-6 Performance Tools V4R2
Bibliography
The following is a list of related printed information that (The commands that are specifically related to
may help you as you use this book. system performance are in Appendix A, “Perform-
ance Tools CL Commands” on page A-1 of this
The books below are listed with their full title and base book.)
order number.
Ÿ AS/400 Licensed Internal Code Diagnostic Aids -
Ÿ BEST/1 Capacity Planning Tool, SC41-5341, pro- Volume 1, LY44-5900, provides the software service
vides information about determining your current representative with an information source about
system performance and predicting your future data error logs, dumps, and traces.
processing needs by using BEST/1 to create a
Ÿ Software Installation, SC41-5120, provides the
model of your system and analyze it. This book
system operator or system administrator with step-
contains scenarios that will help you get started with
by-step procedures for installing the licensed pro-
capacity planning, plus in-depth information about
grams from IBM.
specific topics, such as memory modeling.
Ÿ Work Management, SC41-5306, provides the pro-
Ÿ CL Reference, SC41-5722, provides the application
grammer with information about how to create an
programmer or programmer with a description of the
initial work management environment and how to
AS/400 control language (CL) and its commands.
change it.

 Copyright IBM Corp. 1998 X-1

X-2 Performance Tools V4R2
Index
activity level (continued)
Special Characters MPL (multiprogramming level)
---------- (pgmname) column 7-102 response time B-1
/F indicator 7-115 transaction report (BMPL) 7-107
/H indicator 7-115 multiprogramming level (MPL) 7-46
*DI response time B-1
transaction boundary definitions B-3 transaction report (BMPL) 7-107
*DI value setting 7-19
Transaction Report 8-1 Activity level column 7-103
*DQ Activity Level Time column 7-103
transaction boundary definitions B-3 activity report
*DQ value printing A-43
Transaction Report 8-1 Add Performance Collection (ADDPFRCOL)
*SAVSYS (save system) authority 2-1 command 3-12
>8.0 column 7-102 Add Performance Collection (ADDPFRCOL)
display 3-12
Add Performance Explorer Definition (ADDPEXDFN)
Numerics command 11-6, A-2
0.0-1.0 column 7-144
adding
1.0-2.0 column 7-144
new performance collections 3-12
2.0-4.0 column 7-144
performance explorer definition 11-6, A-2
4.0-8.0 column 7-144
ADDPEXDFN (Add Performance Explorer Definition)
command 11-6, A-2
A ADDPFRCOL (Add Performance Collection)
command 3-12
A-I Wait /Tns column 7-102
Aborts Recd column 7-102 address offset column 11-23
access group advanced peer-to-peer networking (APPN)
analyzing 10-2, A-5 configuration changes 8-14
displaying A-30 control point
access path report printer file 10-13 performance measurements 8-11
accessing work management functions 6-8 control point presentation services (CPPS) 8-15
Act Jobs column 7-102 control point session activation and
Act Level column 7-102 deactivation 8-14
Act Lvl column 7-102 directory services registration and deletion
Act-Inel column 7-102 requests 8-13
Act-Wait column 7-102 local location list updates
Active Devices column 7-102 configuration changes 8-14
active display stations (local or remote) 7-103 remote location list updates
active job configuration changes 8-14
description 6-1 session setup activities 8-16
interactive transactions count 7-40 session setup work activities 8-17
restrictions 6-1 session traffic
Active Jobs column 7-103 QAPMSNA file 8-11
Active Jobs Per Interval column 7-103 work activity
Active K/T /Tns column 7-103 topology maintenance 8-12
active state 7-133 advanced program-to-program communications (APPC)
active task inbound and outbound transactions 8-21
restrictions 6-1 SNA performance measurements 8-3
Active Wrk Stn column 7-103 advisor 4-8
Active/Rsp column 7-103 changing system tuning values 4-10
activity level conclusions, understanding 4-11
definition 1-9 description 4-1

 Copyright IBM Corp. 1998 X-3

advisor (continued) ANZDBF (Analyze Database File) command 10-11,
histogram, using 4-5 A-6
interval conclusions, understanding 4-12 ANZDBFKEY (Analyze Database File Keys)
performance data, collecting 4-2 command 10-13, A-7
requesting analysis 4-3 ANZPFRDTA (Analyze Performance Data)
restrictions 4-1 command 4-3, A-7
selecting a member 4-4 ANZPGM (Analyze Program) command 8-3, 10-9, A-8
selecting time intervals 4-4 APPC (advanced program-to-program communications)
starting 4-3 controller description 8-3
trace data inbound and outbound transactions 8-21
analyzing 4-7 SNA performance measurements 8-3
tune system by recommendations of 4-14 APPN (advanced peer-to-peer networking)
using results 4-7 configuration changes 8-14
Agent feature control point
comparison D-1 performance measurements 8-11
relationship to Manager feature 1-3 control point presentation services (CPPS) 8-15
AI column 11-24 control point session activation and
analysis by interactive key/think time 7-39 deactivation 8-14
analysis by interactive response time 7-39 directory services registration and deletion
analysis by interactive transaction categories 7-39 requests 8-13
Analyze Access Group (ANZACCGRP) local location list updates
command 10-18, A-5 configuration changes 8-14
Analyze BEST/1 Model (ANZBESTMDL) remote location list updates
command A-2 configuration changes 8-14
Analyze Database File (ANZDBF) command 10-9, session setup activities 8-16
10-11, A-6 session setup work activities 8-17
Analyze Database File Keys (ANZDBFKEY) session traffic
command 10-13, A-7 QAPMSNA file 8-11
Analyze Performance Data (ANZPFRDTA) work activity
command 4-3, A-7 topology maintenance 8-12
Analyze Program (ANZPGM) command 8-3, 10-9, A-8 area fill option 9-19
analyzing Arith Ovrflw column 7-103
access group 10-2, A-5 AS/400 Business Graphics Utility (BGU) 9-1
database file 10-2, A-6 ASP ID column 7-103
database file keys 10-13, A-7 assigning
file use 3-16 threads numbers 7-136
job flow 10-2 ASYNC (asynchronous communications)
keys for database files report 10-16 database file 3-1
performance data A-7 Async column 7-103
physical/logical file relationships 10-2 Async DIO /Tns column 7-103
process information 10-16 Async Disk I/O column 7-103
program 8-3, 10-2, A-8 Async Disk I/O per Second column 7-103
program/file use 10-2 Async Disk I/O Requests column 7-103
relationships of programs 10-9 Async I/O /Sec column 7-103
seize/lock conflicts 7-64 Async I/O Per Second column 7-104
system problems, example 13-1 Async Max column 7-104
trace data 4-7 Async Sum column 7-104
ANZACCGRP (Analyze Access Group) asynchronous communications (ASYNC)
command 10-18, A-5 database file 3-1
ANZACCGRP summary asynchronous communications data
environment summary 10-19 database file 3-4
file summary 10-22 Asynchronous DBR column 7-104
job summary 10-20 Asynchronous DBW column 7-104
ANZBESTMDL (Analyze BEST/1 Model) Asynchronous disk I/O
command A-2 definition 7-42

X-4 Performance Tools V4R2

Asynchronous disk I/O per transaction column 7-104 Batch Job Trace Report
Asynchronous NDBR column 7-104 description 7-90
Asynchronous NDBW column 7-104 Job Summary section 7-90
Aut Lookup column 7-105 printing 7-89, 7-90
Authority Lookup sample reports 7-90
caching capability 7-105 Batch permanent writes per second column 7-107
column 7-105 Batch synchronous I/O per second column 7-107
definition 7-105 BBI (begin bracket indicator) bracket delimiter
automatic session traffic fields 8-7
data collection E-1 BCPU / Synchronous DIO column 7-107
performance 3-11 begin bracket indicator (BBI) bracket delimiter
setting up 3-13 session traffic fields 8-7
refresh mode BEST/1 capacity planner
considerations 6-4 agent feature 1-6
restrictions 6-4 introduction 1-6
updating the display 6-4 manager feature 1-6
auxiliary storage management BGU (Business Graphics Utility) 9-1
trace event descriptions 11-26 Bin column 7-107
auxiliary storage pool Binary Overflow column 7-107
definition 4-8 binary synchronous communications (BSC)
Avail Local Storage (K) column 7-105 database file 3-1
Available Storage column 7-105 binary synchronous communications data
Average column 7-105 database file 3-4
Average DIO/Transaction column 7-105 bind command
Average Disk Activity Per Hour column 7-105 definition 8-16
Average K per I/O column 7-105 session setup activities 8-16
Average Phys I/O /Sec column 7-106 BMPL - Cur and Inl column 7-107
Average Reads/Sec column 7-106 boundaries, transaction B-1
Average Response column 7-106 boundary values 7-39
Average Response Time (seconds) column 7-106 broadcast search
Average Response Time column 7-106 definition 8-16
Average Service Time column 7-106 session setup activities 8-16
Average Wait Time column 7-106 broadcast search received
Average Writes/Sec column 7-106 session setup work activity 8-18
Avg CPU /Tns column 7-106 BSC (binary synchronous communications)
Avg K/T /Tns column 7-106 database file 3-1
Avg Length column 7-106 Bundle Writes System column 7-107
Avg Rsp /Tns column 7-106 Bundle Writes User column 7-107
Avg Rsp (Sec) column 7-106 bus counter database file 3-4
Avg Rsp Time column 7-106 Business Graphics Utility (BGU) 9-1
Avg Sec Locks column 7-106 Bytes per Second Received column 7-107
Avg Sec Seizes column 7-106 Bytes per Second Transmitted column 7-107
Avg Time per Service column 7-106 Bytes Recd per Sec column 7-107
Avg Util column 7-106 Bytes Trnsmitd per Sec column 7-107

B C
BASE cache
trace event descriptions 11-24 effect on authority lookups 7-105
Batch asynchronous I/O per second column 7-106 Cache Hit Statistics column 7-108
Batch CPU seconds per I/O column 7-106 call (external) 10-8
Batch CPU Utilization column 7-107 call level column 11-18
Batch impact factor column 7-107 calls made column 11-16
Batch Job Analysis capacity planning
sample report 7-60 description 1-6
section description 7-44

Index X-5
case study, performance analysis 13-1 Collect Data with Menus display 3-8
Category column 7-107 Collect Performance Data display 3-6, 7-11
CEBI (conditional end bracket indicator) bracket delim- collecting
iter data 3-7
session traffic fields 8-7 performance data
central processing unit (CPU) automatically 3-11
See processing unit STRPFRMON (Start Performance Monitor)
central site data collection E-1 command 3-1
Change Functional Area (CHGFCNARA) sample data 3-3
command A-9 trace data
Change Graph Format (CHGGPHFMT) for Lock Report 7-63
command 9-10, A-10 for Transaction Report 7-37
Change Graph Package (CHGGPHPKG) length of time for measurement 3-2
command A-13 storage restrictions 3-2
Change Graph Package display 9-10 collection points 11-3
Change Job Description (CHGJOBD) command 2-1 Collision Detect column 7-108
Change Job Type (CHGJOBTYP) command A-14 command, CL
Change Network Attributes (CHGNETA) Add Performance Collection (ADDPFRCOL) 3-12
command 8-14 Add Performance Explorer Definition
Change Printer File (CHGPRTF) command 2-1 (ADDPEXDFN) 11-6, A-2
Change Program (CHGPGM) command 11-3 ADDPEXDFN (Add Performance Explorer
changing Definition) 11-6, A-2
functional area A-9 ADDPFRCOL (Add Performance Collection) 3-12
graph format 9-10, A-10 Analyze Access Group (ANZACCGRP) A-5
graph package A-13 Analyze BEST/1 Model (ANZBESTMDL)
job description 2-1 command A-2
job type on Print Transaction Report (PRTTNSRPT) Analyze Database File (ANZDBF) A-6
command A-14 Analyze Database File Keys (ANZDBFKEY) 10-13,
network attributes 8-14 A-7
printer file 2-1 Analyze Performance Data (ANZPFRDTA) A-7
program 11-3 Analyze Performance Data (ANZPFRDTA)
system tuning values 4-10 command 4-3
Channel column 7-108 Analyze Program (ANZPGM) 10-9, A-8
checklist, performance and tuning F-1 ANZACCGRP (Analyze Access Group) 10-18, A-5
CHGFCNARA (Change Functional Area) ANZBESTMDL (Analyze BEST/1 Model)
command A-9 command A-2
CHGGPHFMT (Change Graph Format) ANZDBF (Analyze Database File) 10-2, 10-11, A-6
command 9-10, A-10 ANZDBFKEY (Analyze Database File Keys) 10-2,
CHGGPHPKG (Change Graph Package) 10-13, A-7
command A-13 ANZPFRDTA (Analyze Performance Data) A-7
CHGJOBD (Change Job Description) command 2-1 ANZPFRDTA (Analyze Performance Data)
CHGJOBTYP (Change Job Type) command A-14 command 4-3
CHGNETA (Change Network Attributes) ANZPGM (Analyze Program) 10-2, 10-9, A-8
command 8-14 Change Functional Area (CHGFCNARA) A-9
CHGPGM (Change Program) command 11-3 Change Graph Format (CHGGPHFMT) 9-10, A-10
CHGPRTF (Change Printer File) command 2-1 Change Graph Package (CHGGPHPKG) A-13
class-of-service (COS) updates Change Job Description (CHGJOBD) 2-1
configuration changes 8-14 Change Job Type (CHGJOBTYP) A-14
Client Access server Change Network Attributes (CHGNETA) 8-14
job types 7-118, 7-139 Change Printer File (CHGPRTF) 2-1
Cmn column 7-108 Change Program (CHGPGM) 11-3
Cmn I/O column 7-108 CHGFCNARA (Change Functional Area) A-9
Cmn I/O Per Second column 7-108 CHGGPHFMT (Change Graph Format) 9-10, A-10
Collect Additional Data display 3-8, 3-9 CHGGPHPKG (Change Graph Package) A-13
Collect Data with Defaults display 3-7, 7-12 CHGJOBD (Change Job Description) 2-1
CHGJOBTYP (Change Job Type) A-14

X-6 Performance Tools V4R2

command, CL (continued) command, CL (continued)
CHGNETA (Change Network Attributes) 8-14 DLTPFRDTA (Delete Performance Data) 9-14,
CHGPGM (Change Program) 11-3 14-3, A-29
CHGPRTF (Change Printer File) 2-1 DMPTRC (Dump Trace) 3-3
Convert Performance Thread Data DSPACCGRP (Display Access Group) 10-2, 10-16,
(CVTPFRTHD) 15-6 A-30
Copy Functional Area (CPYFCNARA) A-17 DSPAPPNINF (Display APPN Information) 8-13
Copy Graph Format (CPYGPHFMT) 9-11, A-17 DSPFFD (Display File Field Description) 11-8
Copy Graph Package (CPYGPHPKG) 9-11, A-18 DSPHSTGPH (Display Historical Graph) 9-2, A-32
Copy Performance Data (CPYPFRDTA) A-19 DSPPFRDTA (Display Performance Data) 5-1,
CPYFCNARA (Copy Functional Area) A-17 A-35
CPYGPHFMT (Copy Graph Format) 9-11, A-17 DSPPFRGPH (Display Performance Graph) 9-2,
CPYGPHPKG (Copy Graph Package) 9-11, A-18 A-36
CPYPFRDTA (Copy Performance Data) A-19 Dump Trace (DMPTRC) 3-3
Create BEST/1 Model (CRTBESTMDL) End Job Trace (ENDJOBTRC) 10-1, 10-3, A-41
command A-2 End Performance Explorer (ENDPEX) 11-7, A-42
Create Bound C Program (CRTBNDC) 11-3 End Performance Monitor (ENDPFRMON) 3-3
Create Functional Area (CRTFCNARA) A-20 ENDJOBTRC (End Job Trace) 10-1, 10-3, A-41
Create Graph Format (CRTGPHFMT) A-21 ENDPEX (End Performance Explorer) 11-7, A-42
Create Graph Package (CRTGPHPKG) A-24 ENDPFRMON (End Performance Monitor) 3-3
Create Historical Data (CRTHSTDTA) A-25 Print Activity Report (PRTACTRPT) 6-9, A-43
CRTBESTMDL (Create BEST/1 Model) Print Component Report (PRTCPTRPT) 3-5, 7-25,
command A-2 A-45
CRTBNDC (Create Bound C Program) 11-3 Print Job Report (PRTJOBRPT) 7-66, A-48
CRTFCNARA (Create Functional Area) A-20 Print Job Trace (PRTJOBTRC) 10-1, A-51
CRTGPHFMT (Create Graph Format) A-21 Print Lock Report (PRTLCKRPT) 3-5, 7-63, A-52
CRTGPHPKG (Create Graph Package) A-24 Print Performance Explorer Report
CRTHSTDTA (Create Historical Data) A-25 (PRTPEXRPT) 11-8, A-53
CVTPFRTHD (Convert Performance Thread Print Pool Report (PRTPOLRPT) 7-72, A-56
Data) 15-6 Print Resource Report (PRTRSCRPT) 7-76, A-59
Delete BEST/1 Model (DLTBESTMDL) Print System Report (PRTSYSRPT) 3-1, 3-5, A-60
command A-2 Print Trace Report (PRTTRCRPT) 3-5, A-66
Delete Functional Area (DLTFCNARA) A-26 Print Transaction Report (PRTTNSRPT) 3-5, 7-36,
Delete Graph Format (DLTGPHFMT) A-27 7-90, A-64
Delete Graph Package (DLTGPHPKG) A-27 PRTACTRPT (Print Activity Report) 6-9, A-43
Delete Historical Data (DLTHSTDTA) A-28 PRTCPTRPT (Print Component Report) 3-5, 7-25,
Delete Performance Data (DLTPFRDTA) 9-14, A-45
A-29 PRTJOBRPT (Print Job Report) 7-66, A-48
Delete Performance Explorer Data PRTJOBTRC (Print Job Trace) 10-1, 10-4, A-51
(DLTPEXDTA) A-29 PRTLCKRPT (Print Lock Report) 3-5, 7-63, A-52
Display Access Group (DSPACCGRP) 10-2, A-30 PRTPEXRPT (Print Performance Explorer
Display APPN Information (DSPAPPNINF) 8-13 Report) 11-8, A-53
Display File Field Description (DSPFFD) 11-8 PRTPOLRPT (Print Pool Report) 7-72, A-56
Display Historical Graph (DSPHSTGPH) 9-2, A-32 PRTRSCRPT (Print Resource Report) 7-76, A-59
Display Performance Data (DSPPFRDTA) 5-1, PRTSYSRPT (Print System Report) 3-1, 3-5, A-60
A-35 PRTTNSRPT (Print Transaction Report) 3-5, 7-36,
Display Performance Graph (DSPPFRGPH) 9-2, 7-90, A-64
A-36 PRTTRCRPT (Print Trace Report) 3-5, A-66
DLTBESTMDL (Delete BEST/1 Model) Remove Performance Explorer Definition
command A-2 (RMVPEXDFN) A-67
DLTFCNARA (Delete Functional Area) A-26 Rename Object (RNMOBJ) 2-1
DLTGPHFMT (Delete Graph Format) A-27 RMVPEXDFN (Remove Performance Explorer Defi-
DLTGPHPKG (Delete Graph Package) A-27 nition) A-67
DLTHSTDTA (Delete Historical Data) A-28 RNMOBJ (Rename Object) 2-1
DLTPEXDTA (Delete Performance Explorer Start BEST/1 (STRBEST) command A-2
Data) A-29 Start Job Trace (STRJOBTRC) 10-3, A-68

Index X-7
command, CL (continued) Communications Line Detail (continued)
Start Performance Explorer (STRPEX) 11-7, A-69 BSC sample report 7-85
Start Performance Graphics (STRPFRG) 9-2, A-70 DDI sample report 7-84
Start Performance Monitor (STRPFRMON) 3-1, 5-1 ELAN sample report 7-83
Start Performance Tools (STRPFRT) 2-2, A-70 FRLY sample report 7-84
Start Service Job (STRSRVJOB) 10-3 IDLC sample report 7-87
STRBEST (Start BEST/1) command A-2 ISDN Network Interface sample report 7-85
STRJOBTRC (Start Job Trace) 10-1, 10-3, A-68 NWI Maintenance sample report 7-86
STRPEX (Start Performance Explorer) 11-7, A-69 SDLC sample report 7-81
STRPFRG (Start Performance Graphics) 9-2, A-70 section description 7-77
STRPFRMON (Start Performance Monitor) 3-1, 5-1 TRLAN sample report 7-82
STRPFRT (Start Performance Tools) 2-2, A-70 X.25 sample report 7-82
STRSRVJOB (Start Service Job) 10-3 Communications Lines column 7-108
Trace Job (TRCJOB) 10-4 Communications Summary
TRCJOB (Trace Job) 10-4 sample report 7-23
Work with Active Jobs (WRKACTJOB) 2-3 section description 7-19
Work with Disk Status (WRKDSKSTS) 2-3 communications time B-1
Work with Functional Areas (WRKFCNARA) A-71 Component Interval Activity
Work with Job (WRKJOB) 2-3, 6-6 sample report 7-29
Work with Performance Collection section description 7-25
(WRKPFRCOL) 3-12, 9-16, E-1 Component Report
Work with Spooled Files (WRKSPLF) 7-14 Component Interval Activity section 7-25
Work with Submitted Jobs (WRKSBMJOB) 2-3, Database Journaling Summary 7-27
7-13 Database Journaling Summary section 7-27
Work with Subsystems (WRKSBS) 2-3 description 7-25
Work with System Activity (WRKSYSACT) 6-1, 6-2, Disk Activity section 7-26
A-72 Exception Occurrence Summary and Interval Counts
Work with System Status (WRKSYSSTS) 2-3 section 7-27
WRKACTJOB (Work with Active Jobs) 2-3 IOP (input/output processor) Utilizations
WRKDSKSTS (Work with Disk Status) 2-3 section 7-26
WRKFCNARA (Work with Functional Areas) A-71 Job Workload Activity section 7-26
WRKJOB (Work with Job) 2-3, 6-6 Local Work Stations section 7-26
WRKPFRCOL (Work with Performance printing 7-25, A-45
Collection) 3-12, 9-16, E-1 Remote Work Stations section 7-27
WRKSBMJOB (Work with Submitted Jobs) 2-3, Report Selection Criteria section 7-28
7-13 sample reports 7-29
WRKSBS (Work with Subsystems) 2-3 Storage Pool Activity section 7-26
WRKSPLF (Work with Spooled Files) 7-14 composite bar graph 9-6
WRKSYSACT (Work with System Activity) 6-1, 6-2, Concurrent Batch Job Statistics
A-72 sample report 7-61
WRKSYSSTS (Work with System Status) 2-3 section description 7-44
communications controller database file 3-4 conditional end bracket indicator (CEBI) bracket delim-
Communications I/O Count column 7-108 iter
Communications I/O Get column 7-108 session traffic fields 8-7
Communications I/O Put column 7-108 configuration changes
communications interval data APPN local location list updates 8-14
displaying 5-13 APPN remote location list updates 8-14
Communications IOP Utilizations class-of-service (COS) updates 8-14
sample report 7-87 mode updates 8-14
communications line Configure and Manage Tools display 12-1
displaying detail 5-11 configured ASPs column 11-13
displaying interval performance data 5-11 Confirm Create of Historical Data display 9-14, 14-2
performance data time interval 5-13 connection fields
summary of usage 7-19 description 8-4
Communications Line Detail SNA performance measurements 8-4
ASYNC sample report 7-84

X-8 Performance Tools V4R2

control point (CP) capabilities CPU QM column 7-109
control point presentation services (CPPS) 8-15 CPU Sec /Sync DIO column 7-109
control point presentation services (CPPS) CPU Sec Avg and Max column 7-109
APPN (advanced peer-to-peer networking) 8-15 CPU Sec column 7-109
control point (CP) capabilities 8-15 CPU Sec per Tns column 7-109
directory search (DS) 8-16 CPU Seconds column 7-109
registration and deletion 8-16 CPU seconds per transaction column 7-109
topology database update 8-15 CPU Util column 7-109
control point session activation and deactivation CPU Util per Transaction 7-109
APPN (advanced peer-to-peer networking) 8-14 CPU/Async I/O column 7-109
control traffic work load CPU/Sync I/O column 7-109
estimating 8-11 Cpu/Tns (Sec) column 7-109
Control Units column 7-108 CPU/Tns column 7-109
controller description CPYFCNARA (Copy Functional Area) command A-17
APPC protocol 8-3 CPYGPHFMT (Copy Graph Format) command 9-11,
Convert Performance Thread Data (CVTPFRTHD) A-17
command 15-6 CPYGPHPKG (Copy Graph Package) command 9-11,
converting A-18
down-level performance data 12-9, 15-4 CPYPFRDTA (Copy Performance Data)
historical data 12-10 command A-19
performance data Create BEST/1 Model (CRTBESTMDL) command A-2
Agent feature 15-4 Create Bound C Program (CRTBNDC) command 11-3
Manager feature 12-8 Create Functional Area (CRTFCNARA)
thread data 15-6 command A-20
Copy Functional Area (CPYFCNARA) command A-17 Create Graph Format (CRTGPHFMT) command A-21
Copy Graph Format (CPYGPHFMT) command 9-11, Create Graph Package (CRTGPHPKG)
A-17 command A-24
Copy Graph Package (CPYGPHPKG) command 9-11, Create Graph Package display 9-9
A-18 Create Historical Data (CRTHSTDTA) command A-25
Copy Performance Data (CPYPFRDTA) creating
command A-19 functional area A-20
Copy Performance Data Member display 12-7, 15-3 graph format A-21
copying graph package A-24
functional area A-17 historical data A-25
graph format 9-11, A-17 ILE C/400 module 11-3
graph package 9-11, A-18 CRTBESTMDL (Create BEST/1 Model) command A-2
performance data A-19 CRTBNDC (Create Bound C Program) command 11-3
Agent feature 15-2 CRTFCNARA (Create Functional Area)
Manager feature 12-7 command A-20
correlation fields CRTGPHFMT (Create Graph Format) command A-21
description 8-4 CRTGPHPKG (Create Graph Package)
SNA performance measurements 8-4 command A-24
Count column 7-108 CRTHSTDTA (Create Historical Data) command A-25
counting Ctl column 7-110
users 2-1 cum % column 11-18
CPU /Tns column 7-108 Cum CPU Util column 7-110
CPU (central processing unit) Cum Pct Tns column 7-110
See processing unit Cum Util column 7-110
CPU (us) column 11-14 cumulative stats column 11-17
CPU column 7-108 Cur Inl MPL column 7-110
CPU Model column 7-109 Cur MPL column 7-110
CPU per I/O Async column 7-109 CurrentPgm column 11-23
CPU per I/O Sync column 7-109 CVTPFRTHD (Convert Performance Thread Data)
CPU per Logical I/O column 7-109 command 15-6
CPU percent column 11-14

Index X-9
 database file (continued)
D QAPMSNA (SNA performance measurements
DASD Ops per sec column 7-110 data) 3-4
DASD Ops per sec reads column 7-110 QAPMSTND (DDI station counter data) 3-4
DASD Ops per sec writes column 7-110 QAPMSTNY (Frame relay station counter data) 3-4
DASD server QAPMSYS (system data) 3-4
trace event descriptions 11-31 QAPMTSK (task related performance data) 3-4
data (trace) 10-8 QAPMX25 (X.25 data) 3-4
data areas column 11-13 QAPTAZDR (input file) 10-11, 10-13
data collection QAPTLCKD 7-101
automatic 3-11 QAPTLCKD (input file) 7-63
performance data 3-1 QAPTPAGD (input file) 10-18
sample 3-3 QAPTTRCJ (output file) 10-3
setting the time 3-8 QAVPETRCI (collected performance explorer
setup 3-13 data) 11-8
summary 3-13 QPPTSYSR (System Report file) 7-14
system performance data 3-1 QTRDMPT 7-96
system-level analysis 3-13 QTRJOBT 7-91
trace 3-3 QTRJSUM 7-94
using menus 3-8 QTRTSUM 7-91
when to collect 3-1 database file keys
when to end 3-3 analyzing 10-13, A-7
data queue (*DQ) value 8-1 Database Journaling Summary
data types sample report 7-34
valid for X- and Y-axis values 9-7 section description 7-27
database file Database Relation Cross Reference Report 10-12
analyzing A-6 database structure analysis 3-16
high number of file opens and closes 7-41 DB Fault column 7-110
performance trace 7-90 DB Pages column 7-110
QAITMON (collected performance data) 6-1, 6-8 DB Read column 7-110
QAPMASYN (asynchronous data) 3-4 DB Write column 7-110
QAPMBSC (binary synchronous communications DB Wrt column 7-110
data) 3-4 DDM I/O column 7-110
QAPMBUS (bus counter data) 3-4 DDM Svr Wait /Tns column 7-110
QAPMCIOP (communications controller data) 3-4 Dec column 7-111
QAPMCONF (system configuration data) 3-4 Decimal Data column 7-111
QAPMDBMON (Monitoring data) 3-4 Decimal Overflow column 7-111
QAPMDDI (distributed data interface data) 3-4 default
QAPMDIOP (storage device controller data) 3-4 forms size
QAPMDISK (disk unit) 3-4 printer files 2-1
QAPMDMPT (trace database file) 3-5, 7-63 output queue
QAPMECL (establish communications link data) 3-4 performance job description 2-1
QAPMETH (Ethernet statistics data) 3-4 defaults
QAPMFRLY (Frame relay data) 3-4 using to print performance reports 7-11
QAPMHDLC (high-level data link control data) 3-4 defined by column 11-10
QAPMHDWR (hardware configuration data) 3-4 definition description column 11-10
QAPMIOPD (communications processor performance Definition Information 11-9
data) 3-4 definition name column 11-10
QAPMJOBS (job data) 3-4 Delete BEST/1 Model (DLTBESTMDL) command A-2
QAPMLIOP (local work station controller data) 3-4 Delete Functional Area (DLTFCNARA) command A-26
QAPMMIOP (multifunction controller data) 3-4 Delete Graph Format (DLTGPHFMT) command A-27
QAPMPOOL (main storage data) 3-4 Delete Graph Package (DLTGPHPKG) command A-27
QAPMRESP (local work station response time Delete Historical Data (DLTHSTDTA) command A-28
data) 3-4 Delete Performance Data (DLTPFRDTA)
QAPMRWS (remote work station controller command 9-14, 14-3, A-29
data) 3-4

X-10 Performance Tools V4R2

Delete Performance Data display 12-5, 15-1 Disk Feature column 7-111
Delete Performance Explorer Data (DLTPEXDTA) Disk I/O Async column 7-111
command A-29 Disk I/O per Second column 7-112
deleting Disk I/O Reads /Sec column 7-112
functional area A-26 Disk I/O Requests column 7-112
graph format A-27 Disk I/O Sync column 7-112
graph package A-27 Disk I/O Writes /Sec column 7-112
historical data A-28 disk interval
performance data A-29, E-4 displaying 5-11
Agent feature 15-1 Disk IOPs column 7-112
historical data 9-14 Disk mirroring column 7-112
Manager feature 12-5 Disk Space Used column 7-112
performance explorer data A-29 disk status
Description column 7-111, 11-10 working with 2-3
destination network node Disk transfer size (KB) column 7-112
session setup work activity 8-18 disk unit
Detail Activity Report 6-12 displaying 5-11
Detected Access Transmission Error (DTSE) In Disk Utilization
column 7-111 sample report 7-22
Detected Access Transmission Error (DTSE) Out section description 7-19
column 7-111 Disk utilization column 7-112
Device column 7-111 Disk Utilization Detail
device description fields sample report 7-80
description 8-5 section description 7-77
SNA performance measurements 8-5 Disk Utilization Summary
device selection sample report 7-80
definition 8-20 section description 7-77
DIO/Sec Async column 7-111 display
DIO/Sec Sync column 7-111 Add Performance Collection (ADDPFRCOL) 3-12
directed search Change Graph Package 9-10
definition 8-16 Collect Additional Data 3-8
session setup activities 8-16 Collect Data with Defaults 3-7, 7-12
directory search (DS) Collect Data with Menus 3-8
control point presentation services (CPPS) 8-16 Collect Performance Data 3-6, 7-11
directory search processing Configure and Manage Tools 12-1
definition 8-20 Confirm Create of Historical Data 9-14, 14-2
directory services registration and deletion requests Copy Performance Data Member 12-7, 15-3
APPN (advanced peer-to-peer networking) 8-13 Create Graph Package 9-9
disk Delete Performance Data 12-5, 15-1
detail, displaying 5-10 Display by Interval 5-5
interval, displaying 5-11 Display by Job Type 5-4
QAPMJOBS performance monitor file 8-11 Display by Subsystem 5-4
threshold valve 7-26 Display Communications Interval Data 5-14
trace event descriptions 11-26 Display Communications Line Detail 5-11
Disk Activity Display Disk Detail 5-10
sample report 7-31 Display Disk Interval 5-11
section description 7-26 Display Graphs and Packages 9-15
disk activity per hour, average 7-26 Display Job Detail 5-7
Disk Arm Seek Distance column 7-111 Display Jobs 5-6
Disk Arms column 7-111 Display Package Contents 9-12
disk capacity 7-26 Display Performance Data 5-2
Disk Capacity column 7-111 Display Pool Detail 5-8
Disk Controllers column 7-111 Display Pool Interval 5-9
Disk CPU Util column 7-111 Display Remote Jobs 5-12
disk detail IBM Performance Tools 2-2, 7-11
displaying performance data 5-10 Manage Performance Data 15-1

Index X-11
display (continued) Display Performance Data display 5-1
Noninteractive Workload 7-138 Display Performance Graph (DSPPFRGPH)
Performance Tools 2-2, 7-11 command 9-2, A-36
Performance Tools Graphics 9-1 Display Pool Detail display 5-8
Performance Utilities 10-1 Display Pool Interval display 5-9
Print Performance Report 7-6, 7-12 Display Remote Jobs display 5-12
Process Access Group 10-16 displaying
Select Categories for Performance Graphs 9-18 access group A-30
Select Categories for Report 7-8 communication interval data 5-13
Select File and Access Group Utilities 10-2 communications line detail 5-11
Select Graph Format 9-21 database file contents 11-8
Select or Omit Pools 7-9 graph
Select Performance Data Member 9-17 area fill option 9-19
Select Performance Member 5-1, 12-7, 15-2 format options 9-4
Select Sections for Report 7-7 historical 9-20
Select Time Intervals 7-8 output option 9-20
Select Time Intervals to Analyze 4-4 overlays 9-21
Select Time Intervals to Display 5-2 performance 9-16
Select Type of Status 2-3 sample data 9-17
Set Data Collection Time 3-8 types 9-4
Set End Time 3-9 types of data to be graphed 9-6
Set Length of Time to Collect Data 3-9 historical graph A-32
Specify Graph Options 9-18 information type 6-7
Specify Graph Overlay Options 9-22 performance data 5-1, A-35
Specify Report Options 7-10, 7-13 by interval 5-5
Start Collecting Data 3-6, 7-11 by job 5-6
System Activity 6-1 by job type 5-4
Work with All Spooled Files 7-14 by subsystem 5-4
Work with Graph Formats and Packages 9-3 for system resources 5-8
Work with Historical Data 9-13, 14-1 graph overlays 9-21
Work with Job Traces 10-1 how to use 5-1
Work with Performance Collection 3-12 performance graph A-36
Work with System Activity 6-2, 6-3 pool detail 5-8
Display Access Group (DSPACCGRP) pool interval 5-9
command 10-16, A-30 remote jobs 5-12
Display APPN Information (DSPAPPNINF) Distribution of Processing Unit Transactions
command 8-13 sample report 7-52
Display by Interval display 5-5 distribution of transactions
Display by Job Type display 5-4 graphical view 7-40
Display by Subsystem display 5-4 Distribution of Transactions by CPU/Transaction
Display Communications Interval Data display section description 7-40
SDLC (synchronous data link control) 5-14 DLTBESTMDL (Delete BEST/1 Model) command A-2
Display Communications Line Detail display 5-11 DLTFCNARA (Delete Functional Area) command A-26
Display Disk Detail display 5-10 DLTGPHFMT (Delete Graph Format) command A-27
Display Disk Interval display 5-11 DLTGPHPKG (Delete Graph Package) command A-27
Display File Field Description (DSPFFD) DLTHSTDTA (Delete Historical Data) command A-28
command 11-8 DLTPEXDTA (Delete Performance Explorer Data)
Display Graphs and Packages display 9-15 command A-29
Display Historical Graph (DSPHSTGPH) DLTPFRDTA (Delete Performance Data)
command 9-2, A-32 command 9-14, A-29
display I/O (*DI) value 8-1 DMPTRC (Dump Trace) command 3-3
Display Job Detail display 5-7 domain broadcast
Display Jobs display 5-6 definition 8-17
Display Package Contents display 9-12 session setup activities 8-17
Display Performance Data (DSPPFRDTA) down-level performance data, converting 15-4
command 5-1, A-35

X-12 Performance Tools V4R2

DSPACCGRP (Display Access Group) event column 11-23
command 10-16, A-30 Event Wait /Tns column 7-113
DSPAPPNINF (Display APPN Information) EVT column 7-113
command 8-13 example
DSPFFD (Display File Field Description) thread numbers 7-136
command 11-8 Exception Occurrence Summary and Interval Counts
DSPHSTGPH (Display Historical Graph) sample report 7-34
command 9-2, A-32 section description 7-27
DSPPFRDTA (Display Performance Data) Exception Type column 7-113
command 5-1, A-35 Exceptional wait
DSPPFRGPH (Display Performance Graph) column 7-114
command 9-2, A-36 definition 7-114
Dump Trace (DMPTRC) command 3-3 Excp column 7-114
dump trace file 7-96 Excp Wait /Tns column 7-114
duration of trace (us) column 11-12 Excp Wait column 7-114
Excp Wait Sec column 7-114
Excs ACTM /Tns column 7-114
E EXID column 11-24
ECL (establish communications link) 3-4 existence end column 11-14
Elapsed Seconds column 7-112 existence start column 11-14
elapsed time (us) column 11-14 Expert Cache column 7-115
Elapsed Time column 7-112 Exposed AP System Journaled column 7-115
Elapsed Time—Seconds column 7-112 Exposed AP System Not Journaled column 7-115
EM3270 Wait /Tns column 7-112 external response (end user) time
Enable performance collection (ENBPFRCOL) param- definition 1-4
eter 11-3 external response time B-1
ENBPFRCOL (Enable performance collection) param- EXTXHINV (external exception handler) 10-8
eter 11-3 EXTXHRET (call termination) 10-8
End Job Trace (ENDJOBTRC) command 10-1, 10-3,
A-41
End Performance Explorer (ENDPEX) command 11-7, F
A-42 facts, performance F-4
End Performance Monitor (ENDPFRMON) Far End Code Violation column 7-115
command 3-3 file
end point session traffic field 8-6 output 10-4
end-user (external) response time B-1 File and Access Group Utilities commands
ending ANZDBF (Analyze Database File) 10-2
data collection 3-3 DSPACCGRP (Display Access Group) 10-2
job trace 10-1, A-41 File column 7-115
performance explorer 11-7 file use analysis 3-16
performance explorer session A-42 file use and database structure report commands
performance monitor 3-3 summary 3-16
ENDJOBTRC (End Job Trace) command 10-3, A-41 File-to-Program Cross-Reference Report 10-11
ENDPEX (End Performance Explorer) command 11-7, finding
A-42 performance explorer definitions 11-9
ENDPFRMON (End Performance Monitor) floating-bar graph
command 3-3 description 9-6
EORn column 7-113 Flp column 7-115
EOTn column 7-113 Flp Overflow column 7-115
Est Of AWS column 7-113 forms size, default 2-1
establish communications link (ECL) 3-4 Frame Retry column 7-115
Estimated Exposr AP Not Jrnld column 7-113 Frames Received Pct Err column 7-115
Estimated Exposr Curr System column 7-113 Frames Received Total column 7-115
Ethernet statistics database file 3-4 Frames Transmitted Pct Err column 7-115
event (handler) 10-8 Frames Transmitted Total column 7-115

Index X-13
full duplex indicator 7-115 graph (continued)
functional area package (continued)
changing A-9 specifying graph options 9-18
copying A-17 type
creating A-20 composite bar 9-6
deleting A-26 floating-bar 9-6
working with 12-2, A-71 line 9-4
Functional Areas column 7-115 scatter plot 9-5
surface 9-5
graph format
G changing A-10
GDF (graphics data format) file 9-1 copying A-17
GETDR (get direct) operation 10-7 creating A-21
GETKY (get by key) operation 10-7 deleting A-27
GETM (get multiple) operation 10-7 graph package
GETSQ (get sequential) operation 10-7 changing A-13
GO LICPGM menu 2-1 copying A-18
graph creating A-24
format deleting A-27
area fill option 9-19 graphics, performance
changing 9-10 displaying 9-20
creating 9-4 graph format options 9-4
displaying 9-15 graph overlays 9-21
QIBMASYNC 9-2 graph package contents 9-12
QIBMCMNIOP 9-2 graph types 9-4
QIBMCPUTYP 9-2 graphs 9-15
QIBMCPYPTY 9-2 historical graphs 9-20
QIBMDSKARM 9-2 maximum number of legend entries 9-9
QIBMDSKIOP 9-2 packages 9-15
QIBMDSKOCC 9-2 performance graphs 9-16
QIBMLWSIOP 9-2 sample graphs 9-12, 9-17
QIBMMFCIOP 9-2 types of data to be graphed 9-6
QIBMMFDIOP 9-2 historical graphs
QIBMPCTDSK 9-2 changing package 9-10
QIBMPKG 9-3 copying package 9-11
QIBMRSP 9-2 creating formats 9-4
QIBMSYNC 9-2 creating historical data 9-14
QIBMTNS 9-2 creating package 9-9
QIBMTOTDSK 9-2 deleting historical data 9-15
search for formats in the library 9-3 deleting package 9-12
select performance data member 9-17 displaying graphs and packages 9-15
selecting categories for performance searching for data 9-14
graphs 9-18 summary 9-1
specifying graph options 9-18 performance graphs
option changing package 9-10
area fill 9-19 copying package 9-11
output 9-20 creating formats 9-4
overlays 9-21 creating package 9-9
package deleting package 9-12
changing 9-10 displaying 9-16
copying 9-11 displaying graphs and packages 9-15
creating 9-9 output option 9-20
deleting 9-12 summary 9-1
displaying 9-15
QIBMPKG 9-3
searching for packages in the library 9-3
selecting performance data members 9-17

X-14 Performance Tools V4R2

 information type (INFTYPE) parameter 6-7
H INFTYPE (information type) parameter 6-7
half duplex indicator 7-115 initial screening
HDW column 7-116 definition 8-20
high priority session traffic field 8-6 initial topology exchange
High Srv Time column 7-116 topology maintenance 8-13
High Srv Unit column 7-116 inline stats column 11-16
High Util column 7-116 INPACING parameter 8-9
High Util Unit column 7-116 input/output processor (IOP)
High Utilization Disk column 7-116 resource name and model number line
High Utilization Unit column 7-116 column 7-117
hints, tuning F-4 utilization 7-26
histogram 4-5 installing Performance Tools
histogram column 11-18 Agent feature 2-1
historical data GO License Program menu options 2-1
converting 12-10 Manager feature 2-1
creating 9-14, 14-2, A-25 integrated services digital network (ISDN)
deleting 9-15, 14-3, A-28 statistics data file
displaying 9-13 QAPMIDLC, B-channel 3-4
displaying graphs 9-20 QAPMLAPD, D-channel 3-4
historical graph Inter CPU Utilization column 7-116
displaying A-32 Interactive CPU Utilization by 5-Minute Intervals
hit % column 11-18 sample report 7-55
hit cnt column 11-18 section description 7-40
HLL-No column 11-23 Interactive Job Detail
Holder Job Name column 7-116 section description 7-67
Holder Number column 7-116 Interactive Job Detail section
Holder Pool column 7-116 sample report 7-69
Holder Pty column 7-116 Interactive Job Summary
Holder Type column 7-116 sample report 7-68
Holder User Name column 7-116 section description 7-67
Holder’s Job Name column 7-116 interactive key/think time, analysis by 7-39
host (internal) response time B-1 Interactive Program Statistics
host controller description sample report 7-59
SNA performance measurements 8-3 section description 7-42
Interactive Program Transaction Statistics
sample report 7-56
I section description 7-41
I Frames Recd per Sec column 7-116
interactive response time
I Frames Trnsmitd per Sec column 7-116
analysis by 7-39
I/O Wait column 7-116
Job Summary Report 7-40
IEID column 11-24
Interactive Response Time by 5-Minute Intervals
ILE C/400 module
sample report 7-55
creating 11-3
section description 7-40
inbound and outbound transactions 8-22
Interactive Throughput by 5-Minute Intervals
include dependent jobs column 11-10
sample report 7-54
Incoming Calls Pct Retry column 7-116
section description 7-40
Incoming Calls Total column 7-116
interactive transaction averages by job type 7-39
Individual Transaction Statistics
interactive transaction categories, analysis by 7-39
sample report 7-59
Interactive Transactions by 5-Minute Intervals
section description 7-43
sample report 7-54, 7-55
Inel Time A-I/W-I column 7-116
section description 7-40
Inel Wait column 7-116
Interactive Workload
ineligible state 7-133
sample report 7-20
information type
section description 7-18
displaying 6-7

Index X-15
intermediate node on directed search job description
session setup work activity 8-18 changing 2-1
intermediate session traffic field 8-6 Job Interval Report 7-132
intermediate session traffic work load description 7-67
estimating 8-11 Interactive Job Detail section 7-67
internal (host) response time Interactive Job Summary section 7-67
definition 1-4 Noninteractive Job Detail section 7-68
elements B-1 Noninteractive Job Summary section 7-67
internal session-level pacing printing 7-66
description 8-9 Report Selection Criteria section 7-68
excessive waiting 8-9 sample reports 7-68
INPACING and OUTPACING parameters 8-8 Job Maximum A-I column 7-117
INTXHINV (internal exception handler) 10-8 Job Maximum A-W column 7-117
INTXHRET (return from an exception) 10-8 Job Maximum CPU Util column 7-117
INVEXIT (call exit routine) 10-8 Job Maximum Phy I/O column 7-118
IOP (input/output processor) Utilizations Job Maximum Rsp column 7-118
sample report 7-32 Job Maximum Tns column 7-118
section description 7-26 Job Maximum W-I column 7-118
IOP Name column 7-117 Job Name column 7-118, 11-15
IOP Name Network Interface column 7-117 Job Number column 7-118
IOP Name/Line 7-117 Job Pty column 7-118
IOP Name(Model) column 7-117 job report
IOP Processor Util Comm column 7-117 printing A-48
IOP Processor Util DASD column 7-117 job schedule function E-2
IOP Processor Util LWSC column 7-117 Job Set column 7-118
IOP Processor Util Total column 7-117 job states
IOP Util column 7-117 active 7-133
IOP Utilizations ineligible 7-133
section description possible 7-46
communications 7-78 wait 7-133
disk 7-78 Job Statistics
local work station 7-78 printing 7-42
multifunction 7-78 sample report 7-58
ITERM (intervening call termination) 10-8 section description 7-42
ITRMXRSG (resignaling exception) 10-8 Job Summary 7-45
Itv End column 7-117 sample report 7-49
section description 7-38
Job Summary Data
J section description 7-46
Java System Summary Data 7-39
trace event descriptions 11-32 job summary file 7-94
job Job Summary Report
database file 3-1 Batch Job Analysis section 7-44
flow 10-2 Concurrent Batch Job Statistics section 7-44
operational environment 3-16 description 7-38
remote 5-12 Distribution of Transactions by CPU/Transaction
restrictions on active 6-1 section 7-40
trace event descriptions 11-28 including special system information 7-41
tracing 10-4 Individual Transaction Statistics section 7-43
working with 6-6 Interactive CPU Utilization by 5-Minute Intervals
job CPU column 11-15 section 7-40
job creating session column 11-13 Interactive Program Statistics section 7-42
job data Interactive Program Transaction Statistics
merging areas 11-7 section 7-41
separating areas 11-7 Interactive Response Time by 5-Minute Intervals
section 7-40

X-16 Performance Tools V4R2

Job Summary Report (continued)
Interactive Throughput by 5-Minute Intervals K
section 7-40 K per I/O column 7-120
Interactive Transactions by 5-Minute Intervals K/T /Tns Sec column 7-120
section 7-40 KB per I/O Read column 7-120
Job Statistics section 7-42 KB per I/O Write column 7-120
Job Summary section 7-38 Key Fields and Select/Omit Listing Report 10-14
Longest Holders of Seize/Lock Conflicts Key/Think /Tns column 7-120
section 7-44 Key/Think column 7-120
Longest Seize/Lock Conflicts section 7-43 keys, software license 2-1
output (QPSPDJS) 7-37
printing 7-36
Priority-Jobtype-Pool Statistics section 7-42
L
L column 7-120
Report Selection Criteria sample report 7-61
LAPD Pct Frames Recd in Error column 7-120
Report Selection Criteria section 7-45
LAPD Pct Frames Trnsmitd Again column 7-120
RPTTYPE(*SUMMARY) option 7-38
LAPD Total Frames Recd column 7-120
sample report 7-49
LAPD Total Frames Trnsmitd column 7-121
Scatter Diagram section 7-41
Last 4 Programs in Invocation Stack column 7-121
Summary of Seize/Lock Conflicts by Object
legends
section 7-41
description 9-9
System Summary Data sections 7-39
maximum number of entries 9-9
Transaction Significance section 7-40
Length of Wait column 7-121
Job Summary—Batch Job Trace Report
Lgl I/O /Sec column 7-121
sample report 7-90
library
section description 7-90
QGPL 12-11, 15-5
job trace
QPFR 2-1, 2-2
analysis 3-15
QPFRDATA 3-7, 9-2
command
QSYS 10-6
ENDJOBTRC (End Job Trace) 10-1
Library column 7-121, 11-10, 11-11
PRTJOBTRC (Print Job Trace) 10-1
LIC-Pgm--Offset column 11-24
STRJOBTRC (Start Job Trace) 10-1
license keys 2-1
data collection command summary 3-15
Licensed Internal Code
database file 3-4
tasks 6-1
ending 10-1, A-41
Line Count column 7-121
flow 10-2
Line Descriptn column 7-121
printing 10-1, A-51
Line Errors column 7-121
report command summary 3-15
line graph 9-4
starting 10-1, A-68
Line Speed column 7-121
statistics 7-42
line transmission
stopping 10-1
description 8-10
working with 10-1
limitations 8-10
job type
Line Util column 7-121
average resource use per transaction 7-19
LKRL column 7-121
changing on Print Transaction Report (PRTTNSRPT)
LKW column 7-121
command A-14
LKWT column 7-121
Job Type column 7-118
Local End Code Violation column 7-122
Job Workload Activity
Local Not Ready column 7-122
sample report 7-29
local work station controller (WSC) 3-4
section description 7-26
Local work station IOP utilization column 7-122
job/task name column 11-14
Local work station IOPs column 7-122
Jobs column 7-120
local work station response time database file 3-4
jobs in session column 11-13
Local Work Station Response Times
jobs/tasks in session column 11-13
sample report 7-88
journal deposits
section description 7-79
System Total column 7-136
System ToUser column 7-136

Index X-17
Local Work Stations – Response Time Buckets Manager feature
sample report 7-33 comparison D-1
section description 7-26 relationship to Agent feature 1-3
locally started sessions (source) map flag column 11-18
session setup work activity 8-17 mapping
lock 7-41 statements to source code 11-35
analyzing conflicts 7-64 Max Util column 7-123
conflicts by object 7-41 Maximum column 7-123
longest conflicts 7-43 measured response time
longest holders of conflicts 7-44 Job Summary Report 7-41
Lock Conflict column 7-122 medium priority session traffic field 8-6
Lock Report Member column 7-123, 11-10
description 7-64 menus
printing 3-5, 7-63, A-52 GO LICPGM 2-1
resource management 7-63 using to print performance reports 7-6
resource management and trace data 7-63 Merge job data (MRGJOB) parameter 11-7
sample detail report 7-66 merging
sample summary report 7-66 job data 11-7
Seize/Lock Statistics by Time of Day 7-64 MI complex instructions
Seize/Lock Wait Statistics Summary 7-64 trace event descriptions 11-34
trace data 7-63 MI CPLX issued column 11-16
Lock Wait /Tns column 7-122 MI-Pgm--Offset column 11-24
lock-wait 7-38 Minimum column 7-123
Logical column 7-122 missed events due to buffering column 11-13
logical DASDs column 11-13 missed events due to recording column 11-13
Logical Database I/O Other column 7-122 mode updates
Logical Database I/O Read column 7-122 configuration changes 8-14
Logical Database I/O Write column 7-122 module column 11-11
Logical DB I/O column 7-122 monitoring specific jobs 6-5
Logical DB I/O Count column 7-122 MPL (multiprogramming level)
Logical Disk I/O column 7-122 in the activity level 7-107
logical file key report printer file 10-14 on the ineligible queue 7-107
Logical File Listing Report 10-13 transaction report (BMPL) 7-107
Logical I/O /Second column 7-122 transition report 7-46
Logical I/O Per Second column 7-122 MRGJOB (Merge job data) parameter 11-7
Long Wait column 7-123 MRT Max Time column 7-123
Long Wait Lck/Oth column 7-123 multifunction controller database file 3-4
Longest Holders of Seize/Lock Conflicts multiple processors 6-3, 6-12, 7-138
sample report 7-60 multiprogramming level (MPL)
section description 7-44 in the activity level 7-107
Longest Seize/Lock Conflicts on the ineligible queue 7-107
sample report 7-59 transaction report (BMPL) 7-107
section description 7-43 transition report 7-46
Loss of Frame Alignment column 7-123
low priority session traffic field 8-6
N
NAGP column 11-24
M name column 11-10, 11-19, 11-23
MAC Errors column 7-123 Nbr A-I column 7-123
machine object Nbr Evt column 7-123
definition 7-99 Nbr Jobs column 7-123
main storage Nbr Sign offs column 7-123
database file 3-4 Nbr Sign ons column 7-123
Main storage (MB) column 7-123 Nbr Tns column 7-123
Manage Performance Data display 15-1 Nbr W-I column 7-124

X-18 Performance Tools V4R2

NDB Read column 7-124
NDB Write column 7-124 O
NDB Wrt column 7-124 Obj T ST column 11-23
network attributes object
changing 8-14 forms size 2-1
network node processing renaming 2-1
session setup work activity 8-18 Object File column 7-125
network priority session traffic field 8-6 Object Library column 7-125
network transmission priority 8-9 Object Member column 7-125
NNS(DLU) network node Object Name column 7-125, 11-23
definition 8-17 Object RRN column 7-125
session setup activities 8-17 Object Type column 7-125
NNS(OLU) network node obsolete topology entry removal
definition 8-17 topology maintenance 8-13
session setup activities 8-17 OCR (over commitment ratio) 7-126
node congestion updates OMIT parameter 7-45
topology maintenance 8-12 Component Report 7-28
Non-DB Fault column 7-124 Component Report sample 7-35
Non-DB Pages column 7-124 System Report 7-20
Noninteractive Job Detail System Report sample 7-24
sample report 7-70 Omit Parameters column 7-126
section description 7-68 Omit system tasks field 7-10
Noninteractive Job Summary OMTCTL (control units excluded) parameter 7-108
sample report 7-69 OMTFCNARA (functional areas excluded)
section description 7-67 parameter 7-115
Noninteractive Workload OMTJOB (jobs excluded) parameter 7-120
sample report 7-21 OMTLINE (communications lines excluded)
Notices xiii parameter 7-108
NPgs column 11-24 OMTSBS (subsystems excluded) parameter 7-134
Number column 7-124, 11-10 OMTUSRID (users excluded) parameter 7-141
Number I/Os per Second column 7-124 one hop search
Number Jobs column 7-124 definition 8-17
Number Lck Cft column 7-124 Op per Second column 7-126
Number Lck Conflict column 7-124 operational environment of jobs 3-16
Number Locks column 7-124 OPM (original program model) 11-35
Number of batch jobs column 7-124 OPTION(*SS) 7-41
number of brackets ended original program model (OPM) 11-35
tpNEB field 8-7 OS/400
number of brackets started collecting data 1-2
tpNBB field 8-7 monitoring 1-2
number of events column 11-12 tuning 1-2
Number of Jobs column 7-124 OS/400 level column 11-13
number of sessions ended OS/400 Performance Monitor 3-11
tpNSE field 8-7 Other Wait /Tns column 7-126
number of sessions started Outgoing Calls Pct Retry column 7-126
tpNSS field 8-7 Outgoing Calls Total column 7-126
Number Seizes column 7-124 OUTPACING parameter 8-9
Number Sze Cft column 7-124 output
Number Sze Conflict column 7-125 files 10-4
Number Tns column 7-125 option 9-20
Number Traces column 7-125 queue 2-1
Number Transactions column 7-125 Overcommitment ratio column 7-126
overhead removed column 11-16
overlays, graph 9-21

Index X-19
 Pct Tns column 7-127
P percent column 11-23
P (Processor Number) column 11-23 Percent Errored Seconds column 7-128
pacing response Percent Frames Received in Error column 7-128
average amount of time spent waiting 8-8 Percent Full column 7-128
pacing window size 8-8 Percent I Frames Trnsmitd in Error column 7-128
packages, graph Percent Severely Errored Seconds column 7-128
changing 9-10 Percent transactions (dynamic no) column 7-128
copying 9-11 Percent transactions (purge no) column 7-128
creating 9-9 Percent transactions (purge yes) column 7-128
deleting 9-12 Percent Util column 7-128
displaying 9-15 Percentage of Processing Unit by Transaction Catego-
selecting performance data members 9-17 ries
specifying graph options 9-18 sample report 7-53
PAG (process access group) performance
analysis 10-16 analysis 1-7, 13-1
utilities 10-2 checklist F-1
PAG column 7-126 concepts 1-2
PAG Fault column 7-127 correlation of System/36 and AS/400 performance
Page Count column 7-127 parameters C-1
page fault data
trace event descriptions 11-27 collection 3-1
pane size column 11-11 converting 15-4
parameter converting down level 15-4
ENDTNS (end transaction) 10-4 copying 15-2
INFTYPE (information type) 6-7 deleting 15-1, E-4
INPACING parameter 8-9 data collection 3-1
OMIT 7-20 displaying data
OMTCTL (control units excluded) 7-108 by interval 5-5
OMTJOB (jobs excluded) 7-120 by job 5-6
OMTLINE (communications lines excluded) 7-108 by job type 5-4
OMTSBS (subsystems excluded) 7-134 communications line detail 5-11
OMTUSRID (users excluded) 7-141 disk interval 5-11
OMTxxx (data records excluded) 7-132 graph overlays 9-21
OUTPACING parameter 8-9 pool detail 5-8
RPTTYPE (report type) 7-36 pool interval 5-9
SELECT 7-20 STRPFRMON (Start Performance Monitor)
SEQ (sequence) 6-12 command 5-1
SLTCTL (control units included) 7-108 ending data collection 3-3
SLTJOB (jobs included) 7-120 managing in a network E-1
SLTLINE (communications lines included) 7-108 measurement analysis 1-6
SLTSBS (subsystems included) 7-134 objectives 1-3
SLTUSRID (users included) 7-141 stopping data collection 3-3
SLTxxx (data records included) 7-132 transaction 10-2
STRTNS (start transactions) 10-4 why manage 1-1
TITLE (title) 6-10 performance analysis
pass-through transaction path 8-34 overview 1-8
Pct CPU By Categories column 7-127 PRTSYSRPT (Print System Report) command
Pct Data Characters Received in Error column 7-127 overview 1-7
Pct Data Characters Transmitted in Error performance collection
column 7-127 ENBPFRCOL (Enable performance collection) 11-3
Pct Ex-Wt /Rsp column 7-127 pre-defined collection points 11-3
Pct Of Tns Categories column 7-127 setup 3-13
Pct PDUs Received in Error column 7-127 performance data
Pct Poll Retry Time column 7-127 analyzing A-7
converting 12-8

X-20 Performance Tools V4R2

performance data (continued) performance explorer report (continued)
converting obsolete 12-9 column descriptions (continued)
copying 12-7, A-19 IEID 11-24
deleting 12-5, A-29 include dependent jobs 11-10
display by interval 5-5 inline stats 11-16
display by job 5-6 job CPU 11-15
display by job type 5-4 job creating session 11-13
display by subsystem 5-4 job name 11-15
displaying 5-1, A-35 job/task name 11-14
displaying graphs 9-16 jobs in session 11-13
performance explorer jobs/tasks in session 11-13
adding library 11-10, 11-11
definition 11-6, A-2 LIC-Pgm--Offset 11-24
database file logical DASDs 11-13
QAVPETRCI 11-8 map flag 11-18
definition member 11-10
changing A-14 MI CPLX issued 11-16
finding 11-9 MI-Pgm--Offset 11-24
removing A-67 missed events due to buffering 11-13
deleting missed events due to recording 11-13
data A-29 module 11-11
ending 11-7 NAGP 11-24
session A-42 name 11-10, 11-19, 11-23
finding NPgs 11-24
definition 11-9 number 11-10
mapping number of events 11-12
statements to source code 11-35 Obj T ST 11-23
reports object name 11-23
printing 11-8 OS/400 level 11-13
starting 11-7, A-69 overhead removed 11-16
performance explorer report P (Processor Number) 11-23
column descriptions pane size 11-11
address offset 11-23 percent 11-23
AI 11-24 pgm/mod CPU 11-16
call level 11-18 PI 11-24
calls made 11-16 pool 11-14
configured ASPs 11-13 PREFIX 11-24
CPU (us) 11-14 priority 11-14
CPU percent 11-14 procedure 11-11
cum % 11-18 program 11-11, 11-23
cumulative stats 11-17 RC Delta 11-23
CurrentPgm 11-23 run cycles 11-23
data areas 11-13 run time (us) 11-23
defined by 11-10 sample interval (ms) 11-11
definition description 11-10 sector 11-24
definition name 11-10 Seg T ST 11-24
description 11-10 selected jobs 11-10
duration of trace (us) 11-12 selected MI complex instructions 11-10
elapsed time (us) 11-14 selected programs 11-11
event 11-23 selected task names 11-10
EXID 11-24 serial number 11-13
existence end 11-14 session name 11-12
existence start 11-14 sessions since IPL 11-12
histogram 11-18 SKP XCH 11-24
hit % 11-18 span 11-24
hit cnt 11-18 ss.mmm 11-23
HLL-No 11-23 start addr 11-18

Index X-21
performance explorer report (continued) performance graphics (continued)
column descriptions (continued) performance graphs (continued)
start time 11-12 creating formats 9-4
started by user 11-13 creating package 9-9
stmt numb 11-18 deleting package 9-12
stop time 11-12 displaying 9-16
suspend time (us) 11-12 output option 9-20
system model 11-13 selecting categories 9-18
system type 11-13 summary 9-1
target system 11-13 starting A-70
task CPU 11-15 performance history
task ID 11-14, 11-23 collecting sample data 3-1
time stamp 11-22 report 3-1
times called 11-16 performance indicators 8-21
total CPU 11-15, 11-16 performance management
total hits 11-15 using OS/400 1-2
total pages memory 11-13 performance measurement
total raw CPU 11-16 collecting data 1-6
total samples 11-15 producing a system report 1-6
total time 11-12 Performance Measurement and SNADS 8-23
trace wrap count 11-12 performance monitor
type 11-10, 11-11 ending 3-3
unit 11-24 job schedule function E-2
unknown CPU 11-16 starting 3-1
user 11-10 performance objectives
version 11-13 establishing 1-3
printing A-53 performance parameters
performance facts F-4 correlation of System/36 and AS/400 C-1
performance graph description C-1
displaying A-36 performance reports
performance graphics anatomy of 7-1
displaying available 7-3
graph format options 9-4 choosing 7-5
graph overlays 9-21 column descriptions 7-102
graph package contents 9-12 Component Report 7-25
graph types 9-4 general description 7-1
graphs 9-15 header information 7-1
historical graphs 9-20 Job Interval Report 7-67
legends, maximum number of entries 9-9 printing 7-6, 7-11
packages 9-15 Transaction Report 7-46
performance graphs 9-16 using defaults to print 7-11
sample graphs 9-12, 9-17 using menus to print 7-6
types of data to be graphed 9-6 Performance Tools
historical graphs Agent feature 1-3
changing package 9-10 and OS/400 function 1-2
copying package 9-11 default output queue 2-1
creating formats 9-4 installing 2-1
creating historical data 9-14 introduction 1-1
creating package 9-9 Manager feature 1-3
deleting historical data 9-15 restrictions 1-3
deleting package 9-12 starting A-70
displaying 9-15, 9-20 Performance Tools Graphics display 9-1
searching for data 9-14 Performance Tools menu 2-2, 7-11
summary 9-1 Performance Utilities display 10-1
performance graphs Perm Write column 7-128
changing package 9-10
copying package 9-11

X-22 Performance Tools V4R2

Permanent writes per transaction column 7-128 Print Trace Report (PRTTRCRPT) command 3-5, A-66
pgm/mod CPU column 11-16 Print Transaction Report (PRTTNSRPT)
Physical I/O Count column 7-128 command 3-5, 7-36, 7-90, A-64
PI (pool identifier) column 11-24 printer file 10-11
Pl (Pool) column 7-129 changing 2-1
planning for performance tuning F-1 characteristics 2-1
planning, capacity 1-6 default forms size 2-1
pool QAPTPAGD (input file) 10-16
detail, displaying 5-8 QPPTANKM 10-14, 10-15
interval, displaying 5-9 QPPTANZD
Pool Activity physical-to-logical database file 10-11
sample report 7-74 QPPTANZK 10-13
section description 7-73 QPPTANZP 10-9
Pool column 7-129, 11-14 QPPTLCK 7-63
Pool ID column 7-129 QPPTPAGD 10-16
Pool ID Faults column 7-129 QPPTTRC1 10-4
Pool Interval Report QPPTTRC2 10-4
description 7-73 QPPTTRCD 10-4
Pool Activity 7-73 QPSPDJS 7-37
printing 7-73 QPSPDTD 7-37
Report Selection Criteria section 7-73 QPSPDTS 7-37
sample reports 7-73 Printer Lines column 7-129
Subsystem Activity section 7-73 Printer Pages column 7-129
Pool Mch Faults/Sec column 7-129 printing
pool report activity report A-43
printing A-56 Batch Job Trace Report 7-90
Pool size (KB) column 7-129 Component Report 7-25, A-45
Pool User Faults/Sec column 7-129 Job Interval Report 7-66
Pools column 7-129 job report A-48
pre-defined collection points 11-3 Job Summary Report 7-36
PREFIX column 11-24 job trace 10-1, A-51
Prg column 7-129 Lock Report 7-63, A-52
Print Activity Report (PRTACTRPT) command 6-9, performance explorer report 11-8, A-53
A-43 performance reports
Print Component Report (PRTCPTRPT) using commands 7-5
command 3-5, 7-25, A-45 using defaults 7-11
Print Job Report (PRTJOBRPT) command 7-66, A-48 using menus 7-6, 7-11
Print Job Trace (PRTJOBTRC) command 10-1, 10-4, Pool Interval Report 7-73
A-51 pool report A-56
Print Lock Report (PRTLCKRPT) command A-52 Resource Interval Report 7-76
considerations 7-63 resource report A-59
lock conflicts 7-63 System Report 7-17, A-60
seize conflicts 7-63 trace report A-66
viewing trace data 3-5 transaction report 3-5, 7-36, 7-90, A-64
Print Performance Explorer Report (PRTPEXRPT) Transition Report 7-36
command 11-8 Priority column 7-129, 11-14
Print Performance Explorer Report Priority-Jobtype-Pool Statistics
(PRTPEXRPT) A-53 sample report 7-58
PRTPEXRPT (Print Performance Explorer Priority-Jobtype-Pool Statistics section
Report) A-53 Job Summary Report 7-42
Print Performance Report display 7-6, 7-12 problem analysis example 13-1
Print Pool Report (PRTPOLRPT) command 7-72, A-56 procedure column 11-11
Print Resource Report (PRTRSCRPT) command 7-76, process
A-59 trace event descriptions 11-28
Print System Report (PRTSYSRPT) command 3-1, process access group (PAG)
3-5, A-60 analysis 10-16

Index X-23
process access group (PAG) (continued) PRTTNSRPT (Print Transaction Report)
utilities 10-2 command 3-5, 7-36, 7-90, A-64
Process Access Group display 10-16 PRTTRCRPT (Print Trace Report) command 3-5, A-66
Process Access Group Information Report 10-17 Pty column 7-130
process data collection command Purge column 7-130
summary 3-17 PUT (add a record) operation 10-7
process data collection report command PUTM (add a record) operation 10-7
summary 3-17 PWrt column 7-130
process information, analyzing 10-16
processing unit
graphical view 7-40 Q
processing time 10-5 QAITMON database file 6-1, 6-8
processing unit 6-3 QAOMJOBS performance monitor file 8-11
trace entry 10-9 QAPMAPPN database file 3-4
utilization, Job Summary Report 7-40 QAPMASYN database file 3-4
processor QAPMBSC database file 3-4
summary of usage 7-19 QAPMBUS database file 3-4
Profile CPU Summary Information Report 11-15 QAPMCIOP database file 3-4
Profile Information Report 11-18, 11-20 QAPMCONF database file 3-4
program QAPMDBMON database file 3-4
analyzing A-8 QAPMDDI database file 3-4
changing 11-3 QAPMDIOP database file 3-4
QCRMAIN 10-6 QAPMDISK database file 3-4
QRGXINIT 10-6 QAPMDMPT database file 3-5, 7-63
QWSGET B-3 QAPMECL database file 3-4
program bracket QAPMETH database file 3-4
trace event descriptions 11-31 QAPMFRLY database file 3-4
Program column 7-129, 11-11, 11-23 QAPMHDLC database file 3-4
program environment 3-16 QAPMHDWR database file 3-4
Program Name column 7-129 QAPMIDLC database file 3-4
Program-to-File Cross-Reference Report 10-10 QAPMIOPD database file 3-4
program-to-file relationship report 10-9 QAPMJOBS database file 3-4
program-to-file relationships, analyzing 10-9 QAPMLAPD database file 3-4
Protocol column 7-130 QAPMLIOP database file 3-4
PRTACTRPT (Print Activity Report) command 6-9, QAPMMIOP database file 3-4
A-43 QAPMPOOL database file 3-4
PRTCPTRPT (Print Component Report) QAPMRESP database file 3-4
command 3-5, 7-25, A-45 QAPMRWS database file 3-4
PRTJOBRPT (Print Job Report) command 7-66, A-48 QAPMSAP database file 3-4
PRTJOBTRC (Print Job Trace) command 10-1, A-51 QAPMSBSD subsystem 3-4
PRTLCKRPT (Print Lock Report) command A-52 QAPMSNA database file 3-4, 8-3
considerations 7-63 QAPMSNADS database file 3-4
lock conflicts 7-63 QAPMSTND database file 3-4
seize conflicts 7-63 QAPMSTNE database file 3-4
viewing trace data 3-5 QAPMSTNL database file 3-4
PRTPEXRPT (Print Performance Explorer Report) QAPMSTNY database file 3-4
command 11-8 QAPMSYS database file 3-4
Print Performance Explorer Report QAPMTSK database file 3-4
(PRTPEXRPT) A-53 QAPMX25 database file 3-4
PRTPEXRPT (Print Performance Explorer QAPTAZDR database file 10-13
Report) A-53 QAPTLCKD database file 7-63
PRTPOLRPT (Print Pool Report) command 7-72, A-56 QAPTLCKD file 7-101
PRTRSCRPT (Print Resource Report) command 7-76, QAPTPAGD database file 10-16, 10-18
A-59 QAPTTRCJ database file 10-3
PRTSYSRPT (Print System Report) command 3-1, QAVPETRCI database file 11-8
3-5, A-60

X-24 Performance Tools V4R2

QBASE subsystem 3-13 received TDUs
QCRMAIN program 10-6 topology maintenance 8-13
QCTL subsystem 3-13 receiver of search requests
QDBPUT database module 10-7 session setup work activity 8-17
QGPL library 12-11, 15-5 receiving performance data 8-10
QIBMASYNC graph format 9-2 recommendations, understanding 4-8
QIBMCMNIOP graph format 9-2 Record Number column 7-130
QIBMCPUTYP graph format 9-2 record selection report printer file 10-13
QIBMCPYPTY graph format 9-2 reducing intermediate session work load 8-11
QIBMDSKARM graph format 9-2 refresh mode
QIBMDSKIOP graph format 9-2 automatic 6-4
QIBMDSKOCC graph format 9-2 registration and deletion
QIBMLWSIOP graph format 9-2 control point presentation services (CPPS) 8-16
QIBMMFCIOP graph format 9-2 remote job
QIBMMFDIOP graph format 9-2 displaying performance data 5-12
QIBMPCTDSK graph format 9-2 Remote LAN Pct Frames Recd column 7-130
QIBMPKG graph package 9-3 Remote LAN Pct Frames Trnsmitd column 7-130
QIBMRSP graph format 9-2 Remote Not Ready column 7-131
QIBMSYNC graph format 9-2 Remote Seq Error column 7-131
QIBMTNS graph format 9-2 Remote Work Station Response Times
QIBMTOTDSK graph format 9-2 sample report 7-89
QPFR library 2-1, 2-2 section description 7-79
QPFRDATA library 3-7, 9-2 Remote Work Stations
QPITACTR spooled file 6-9 section description 7-27
QPPTANKM printer file 10-14, 10-15 Remote Work Stations – Response Time Buckets
QPPTANZD sample report 7-33
database file 10-11 Remove Performance Explorer Definition
printer file 10-11 (RMVPEXDFN) command A-67
QPPTANZK printer file 10-13 removing
QPPTANZP printer file 10-9 performance data E-4
QPPTLCK printer file 7-63 performance explorer data A-29
QPPTPAGD printer file 10-16 performance explorer definition A-67
QPPTSYSR database file 7-14 Rename Object (RNMOBJ) command 2-1
QPPTTRC1 printer file 10-4 renaming
QPPTTRC2 printer file 10-4 object 2-1
QPPTTRCD printer file 10-4 report
QPSPDJS (Job Summary Report output) 7-37 Analysis of Keys for Database Files 10-16
QPSPDTD (Transition Report output) 7-37 ANZACCGRP Summary Report
QPSPDTS (Transaction Report output) 7-37 environment summary 10-19
QRGZINIT program 10-6 file summary 10-22
QSYS library 10-6 job summary 10-20
QTRDMPT file 7-96 Batch Job Trace Report
QTRJOBT file 7-91 description 7-90
QTRJSUM file 7-94 Job Summary 7-90
QTRTSUM file 7-91 Job Summary section 7-90
Queue Length column 7-130 Component Report
QWSGET program B-3 Report Selection Criteria 7-28
Database Relation Cross Reference Report 10-12
Definition Information 11-9
R Detail Activity Report 6-12
Rank column 7-130 File to Program Cross Reference 10-11
Ratio of write disk I/O to total disk I/O column 7-130 Job Interval Report 7-67
RC Delta column 11-23 Job Summary Report
Reads per Second column 7-130 Concurrent Batch Job Statistics section 7-44
Receive CRC Errors column 7-130 Individual Transaction Statistics section 7-43
Longest Holders of Seize/Lock Conflicts
section 7-44

Index X-25
report (continued) Requestor’s Job Name column 7-131
Job Summary Report (continued) Reset Packets Recd column 7-131
Report Selection Criteria 7-45 Reset Packets Trnsmitd column 7-131
Key Fields and Select/Omit Listing 10-14 Resource Interval Report
Lock Report 7-63 Communications Line Detail section 7-77
Logical File Listing 10-13 description 7-76
Pool Interval Report 7-73 Disk Utilization Detail section 7-77
Print Activity 6-9 Disk Utilization Summary section 7-77
Process Access Group Information 10-17 IOP Utilizations section 7-78
Profile CPU Summary Information 11-15 Local Work Station Response Times section 7-79
Profile Information 11-18, 11-20 printing 7-76
Program-to-File Cross-Reference 10-10 Remote Work Station Response Times
Resource Interval Report 7-76 section 7-79
Run Information 11-11 sample reports 7-80
Statistics CPU Summary Information 11-15 resource report
Summary Activity 6-10 printing A-59
System Report 7-17 Resource Utilization
Communications Summary 7-19 sample report 7-21, 7-22
Disk Utilization section 7-19 section description 7-19
Report Selection Criteria 7-20 Resource Utilization Expansion
Resource Utilization 7-19 sample report 7-21
Resource Utilization Expansion 7-19 section description 7-19
Storage Pool Utilization section 7-19 resources, system
Workload 7-18 auxiliary storage 3-1
Task Information 11-13 communications 3-1
Trace Analysis I/O Summary Report 10-6 main storage 3-1
Trace Analysis Summary Report 10-5 processing unit 3-1
Trace Job Information Report 10-7 Response column 7-131
Transaction Report 7-37 Response Sec Avg and Max column 7-131
transaction response, differences in B-2 Response Seconds column 7-131
Transition Report response time
sample 7-46 components B-1
summary 7-46 description 1-4
report command host (internal) B-1
Analyze Access Group (ANZACCGRP) 10-2 input communications time B-1
ANZACCGRP (Analyze Access Group) 10-18 internal 1-4
ANZDBF (Analyze Database File) 10-11 RH (request header) bracket delimiter
ANZPGM (Analyze Program) 10-2, 10-9 session traffic fields 8-7
Display Access Group (DSPACCGRP) 10-2 RMVPEXDFN (Remove Performance Explorer Defi-
DSPACCGRP (Display Access Group) 10-16 nition) command A-67
summary 3-13 RNMOBJ (Rename Object) command 2-1
Report Selection Criteria route selection
Component Report description 7-28 definition 8-20
Component Report sample 7-35 RPTTYPE (report type) parameter 7-36
Job Interval Report 7-68 Rsp column 7-131
Job Interval sample 7-71 Rsp Time column 7-131
Job Summary description 7-45 Rsp Timer Ended column 7-131
Pool Interval Report 7-73 Rsp/Tns column 7-131
Pool Interval sample 7-75 run cycles column 11-23
System Report description 7-20 Run Information Report 11-11
System Report sample 7-24 run time (us) column 11-23
report type (RPTTYPE) parameter 7-36
request header (RH) bracket delimiter
session traffic fields 8-7 S
requesting a performance data analysis 4-3 S/L column 7-131

X-26 Performance Tools V4R2

SACPNM correlation field sample report (continued)
description 8-4 Local Work Stations – Response Time
sample Buckets 7-33
data collection 3-5 Lock Report – Summary 7-66
sample data Lock Report–Detail 7-66
definition 3-3 Longest Holders of Seize/Lock Conflicts 7-60
how to collect 3-3 Longest Seize/Lock Conflicts 7-59
SNADS 8-29 Noninteractive Job Detail 7-70
sample interval (ms) column 11-11 Noninteractive Job Summary 7-69
sample report Noninteractive Workload 7-21
Batch Job Analysis 7-60 performance explorer
Communications IOP Utilizations 7-87 Definition Information 11-9
Communications Line Detail–ANYNC 7-84 Profile CPU Summary Information 11-15
Communications Line Detail–BSC 7-85 Profile Information 11-18, 11-20
Communications Line Detail–DDI 7-84 Run Information 11-11
Communications Line Detail–ELAN 7-83 Statistics CPU Summary Information
Communications Line Detail–FRLY 7-84 Report 11-15
Communications Line Detail–IDLC 7-87 Task Information Report 11-13
Communications Line Detail–ISDN 7-85 Trace Information 11-21
Communications Line Detail–NWI Pool Activity 7-74
Maintenance 7-86 Priority-Jobtype-Pool Statistics 7-58
Communications Line Detail–SDLC 7-81 Remote Work Station Response Times 7-89
Communications Line Detail–TRLAN 7-82 Remote Work Stations – Response Time
Communications Line Detail–X.25 7-82 Buckets 7-33
Communications Summary 7-23 Report Selection Criteria—Component Report 7-35
Component Interval Activity 7-29 Report Selection Criteria–Job Interval 7-71
Concurrent Batch Job Statistics 7-61 Report Selection Criteria–Job Summary
Database Journaling Summary 7-34 Report 7-61
Detail Activity Report 6-12 Report Selection Criteria–Pool Interval 7-75
Disk Activity 7-31 Report Selection Criteria–System Report 7-24
Disk Utilization 7-22 Resource Utilization 7-21
Disk Utilization Detail 7-80 Resource Utilization Expansion 7-21
Disk Utilization Summary 7-80 Storage Pool Activity 7-31
Distribution of Processing Unit Transactions 7-52 Storage Pool Utilization 7-22
Exception Occurrence Summary and Interval Subsystem Activity 7-74
Counts 7-34 Summary Activity Report 6-10
Individual Transaction Statistics 7-59 Summary of Seize/Lock Conflicts by Object 7-57
Interactive CPU Utilization by 5-Minute System Summary Data 7-51, 7-52
Intervals 7-55 Transaction Report 7-62
Interactive Job Detail section 7-69 Transaction Significance 7-53
Interactive Job Summary 7-68 Transition Report 7-62
Interactive Program Statistics 7-56, 7-59 SANWID correlation field
Interactive Response Time by 5-Minute description 8-4
Intervals 7-55 SAPPN correlation field
Interactive Throughput by 5-Minute Intervals 7-54 description 8-4
Interactive Transactions by 5-Minute Intervals 7-54, save system (*SAVSYS) authority 2-1
7-55 Scatter Diagram section 7-41
Interactive Workload 7-20 scatter plot 9-5
IOP (input/output processor) Utilizations 7-32 scenario, performance analysis 13-1
Job Statistics 7-58 SCTLNM correlation field
Job Summary 7-49 description 8-4
Job Summary Report: Report Selection SCTYP correlation field
Criteria 7-61 description 8-4
Job Summary—Batch Job Trace Report 7-90 search processing
Job Workload Activity 7-29 session setup work activity 8-17
Local Work Station Reponse Times 7-88

Index X-27
sector column 11-24 session setup activities
Seg T ST column 11-24 APPN (advanced peer-to-peer networking) 8-16
segment address register bind command 8-16
trace event descriptions 11-28 broadcast search 8-16
Seize and Lock Conflicts column 7-131 directed search 8-16
Seize Conflict column 7-131 domain broadcast 8-17
Seize Hold Time column 7-132 NNS(DLU) network node 8-17
seize lock NNS(OLU) network node 8-17
trace event descriptions 11-28 session setup activities 8-17
Seize Wait /Tns column 7-132 session setup work activity
seize/lock APPN (advanced peer-to-peer networking) 8-17
analyzing conflicts 7-64 broadcast search received 8-18
conflicts by object 7-41 destination network node 8-18
longest conflicts 7-43 device selection 8-20
longest holders of conflicts 7-44 directory search processing 8-20
Seize/Lock Conflicts by Object 7-41 initial screening 8-20
seize/lock data file 7-101 intermediate node on directed search 8-18
Select Categories for Performance Graphs locally started sessions (source) 8-17
display 9-18 network node processing 8-18
Select Categories for Report display 7-8 receiver of search requests 8-17
Select File and Access Group Utilities display 10-2 route selection 8-20
Select Graph Format display 9-21 search processing 8-17
Select or Omit Pools display 7-9 switched link activation 8-20
SELECT parameter session traffic fields
Component Report 7-28 description 8-6
Component Report sample 7-35 number of sessions started and ended 8-7
Job Summary Report 7-45 SNA performance measurements 8-6
System Report 7-20 session-level pacing
System Report sample 7-24 description 8-7
Select Parameters column 7-132 excessive waiting 8-9
Select Performance Data Member display 9-17 sessions since IPL column 11-12
Select Performance Member display 5-1, 12-7, 15-2 Set Data Collection Time display 3-8
Select Sections for Report display 7-7 Set End Time display 3-9
Select Time Intervals display 7-8 Set Length of Time to Collect Data display 3-9
Select Time Intervals to Analyze display 4-4 setting
Select Time Intervals to Display display 5-2 data collection time 3-8
Select Type of Status display 2-3 Short Frame Errors column 7-132
selected jobs column 11-10 Short Wait /Tns column 7-132
selected MI complex instructions column 11-10 Short WaitX /Tns (Short wait extended) column 7-132
selected programs column 11-11 single-processor system
selected task names column 11-10 field restrictions 6-3
selecting Size (K) column 7-132
a member 4-4 Size (M) column 7-132
performance data members 9-17 Size column 7-132
time intervals 4-4 SKP XCH column 11-24
sending data SLINNM correlation field
internal session-level pacing 8-7 description 8-4
line transmission 8-10 SLIOMT correlation field
session-level pacing 8-7 description 8-4
transmission priority 8-7 SLTCTL (control units included) parameter 7-108
SEQ (sequence) parameter 6-12 SLTFCNARA (functional areas included)
sequence (SEQ) parameter 6-12 parameter 7-115
Sequence Error column 7-132 SLTJOB (jobs included) parameter 7-120
serial number column 11-13 SLTLINE (communications lines included)
session name column 11-12 parameter 7-108

X-28 Performance Tools V4R2

SLTSBS (subsystems included) parameter 7-134 start addr column 11-18
SLTUSRID (users included) parameter 7-141 Start BEST/1 (STRBEST) command A-2
SLTxxx (select) parameter 7-132 Start Collecting Data display 3-6, 7-11
SMAPP ReTune column 7-132 Start column 7-133
SNA performance measurements start database monitor (STRDBMON) parameter 3-4
APPC controller description 8-3 Start Job Trace (STRJOBTRC) command 10-1, 10-3,
comparing different traffic priority levels 8-11 A-68
connection fields 8-4 Start Performance Explorer (STRPEX) command 11-7,
correlation fields 8-4 A-69
device description fields 8-5 Start Performance Graphics (STRPFRG)
host controller description 8-3 command 9-2, A-70
QAPMSNA database file 8-3 Start Performance Monitor (STRPFRMON)
session traffic fields 8-6 command 3-1, 5-1
T2 station I/O manager task fields 8-5 Start Performance Tools (STRPFRT) command 2-2,
SNA performance measurements data A-70
database file 3-4 Start Service Job (STRSRVJOB) command 10-3
SNACVO device description field start time column 11-12
description 8-5 started by user column 11-13
SNADD device description field Started column 7-133
description 8-5 starting
SNADS advisor 4-3
performance measurement 8-23 job trace 10-1, 10-3, A-68
performance notes 8-32 performance explorer 11-7, A-69
sample data 8-29 performance graphics 9-2, A-70
SNADS transaction performance monitor 3-1, 5-1
gateway senders 8-28 Performance Tools 2-1, 2-2, A-70
SNADS receiver 8-25 State column 7-133
SNADS router 8-24 State Transitions A-A column 7-133
SNADS sender 8-25 State Transitions A-I column 7-133
SVDS receiver 8-26 Statistics CPU Summary Information Report 11-15
SVDS sender 8-27 STLLBU connection field
SNLBU connection field description 8-5
description 8-5 stmt numb column 11-18
SNWAIN T2 station I/O manager task field Stop column 7-133
description 8-5 stop time column 11-12
SNWAOU T2 station I/O manager task field Stopped column 7-133
description 8-5 stopping
software license keys 2-1 job trace 10-1
SOTn column 7-132 storage device controller database file 3-4
source code storage pool
mapping with statements 11-35 displaying 5-8
span column 11-24 displaying performance data 5-9
Specify Graph Options display 9-18 setting size 7-19
Specify Graph Overlay Options display 9-22 Storage Pool Activity
Specify Report Options display 7-10, 7-13 sample report 7-31
Spool CPU seconds per I/O column 7-133 section description 7-26
Spool database reads per second column 7-133 Storage Pool Utilization
Spool I/O per second column 7-133 sample report 7-22
spooled file section description 7-19
QPITACTR 6-9 STRBEST (Start BEST/1) command A-2
Srv Time column 7-133 STRDBMON (start database monitor) parameter 3-4
ss.mmm column 11-23 STRJOBTRC (Start Job Trace) command 10-3, A-68
SSPTRC (trace suspended) 10-8 STRPEX (Start Performance Explorer) command 11-7,
STACVO device description field A-69
description 8-5 STRPFRG (Start Performance Graphics)
command 9-2, A-70

Index X-29
STRPFRMON (Start Performance Monitor) system (continued)
command 3-1, 5-1 analysis 13-1
STRPFRT (Start Performance Tools) command 2-2, configuration database file 3-4
A-70 database file 3-4
STRSRVJOB (Start Service Job) command 10-3 performance parameters C-1
STSKNM correlation field two processors
description 8-4 example display 7-138
submitted job system activity
working with 2-3 working with 6-1, A-72
subsystem System Activity menu 6-1
database file, QAPMSBSD 3-4 System CPU per transaction (seconds) column 7-136
QBASE 3-13 system data collection command
QCTL 3-13 summary 3-13
working with 2-3 system default
Subsystem Activity collecting data 3-7
sample report 7-74 System disk I/O per transaction column 7-136
section description 7-73 system model column 11-13
Subsystem Name column 7-133 System Report
Subsystems column 7-134 Communications Summary section 7-19
Sum column 7-134 description 7-17
Summary Activity Report 6-10 Disk Utilization section 7-19
Summary of Seize/Lock Conflicts by Object printing 3-1, 7-17, A-60
sample report 7-57 Report Selection Criteria section 7-20
section description 7-41 Resource Utilization Expansion section 7-19
surface graph 9-5 Resource Utilization section 7-19
suspend time (us) column 11-12 sample reports 7-20
SVDS receiver 8-26 Storage Pool Utilization section 7-19
SVDS sender 8-27 Workload section 7-18
switched link activation system resource utilization
definition 8-20 BEST/1 1-6
SWX column 7-134 system resources
Sync column 7-134 auxiliary storage 3-1
Sync DIO /Tns column 7-134 communications 3-1
Sync Disk I/O column 7-134 displaying performance data 5-8
Sync Disk I/O per Second column 7-134 main storage 3-1
Sync Disk I/O Requests column 7-134 processing unit 3-1
Sync Disk I/O Rqs/Tns column 7-134 System Starts column 7-136
Sync I/O /Elp Sec column 7-134 System Stops column 7-136
Sync I/O /Sec column 7-134 System Summary Data
Sync I/O Per Second column 7-134 interactive key/think time, analysis by 7-39
Synchronous DBR column 7-135 interactive response time, analysis by 7-39
Synchronous DBW column 7-135 interactive transaction averages by job type 7-39
Synchronous DIO / Act Sec column 7-135 interactive transaction categories, analysis by 7-39
Synchronous DIO / Ded Sec column 7-135 section description 7-39
Synchronous DIO / Elp Sec column 7-135 trace periods for trace data 7-39
Synchronous disk I/O System Summary Data section
definition 7-42 exceptional wait breakdown by job type 7-39
Synchronous Disk I/O Counts column 7-135 System Total column 7-136
Synchronous disk I/O per transaction column 7-135 System ToUser column 7-136
Synchronous Max column 7-135 system type column 11-13
Synchronous NDBR column 7-136 system use analysis 3-2
Synchronous NDBW column 7-136 system with two processors
Synchronous Sum column 7-136 definition 6-3
Synchronous wrt column 7-136 example display 6-3
system utilization value 6-12
activity 6-1

X-30 Performance Tools V4R2

system-level Total CPU Sec /Sync DIO column 7-137
analysis 3-13 Total CPU Utilization column 7-138
data collection 3-13 Total Data Characters Received column 7-138
System/36 performance parameters Total Data Characters Transmitted column 7-138
AS/400 correlation C-1 Total fields per transaction column 7-138
SZRL column 7-136 Total Frames Recd column 7-138
SZWT column 7-136 total hits column 11-15
Total I Frames Trnsmitd column 7-138
Total I/O column 7-138
T total pages memory column 11-13
T2 station I/O manager (IOM) task fields Total PDUs Received column 7-138
description 8-5 Total Physical I/O per Second column 7-138
SNA performance measurements 8-5 total processing unit (TCPU) seconds
target system column 11-13 Job Summary Report 7-42
task CPU column 11-15 total raw CPU column 11-16
task ID column 11-14, 11-23 Total Responses column 7-138
Task Information Report 11-13 total samples column 11-15
Thread column 7-136 Total Seize/Wait Time column 7-138
thread data total time column 11-12
converting 15-6 Total Tns column 7-138
threads tpIPNW internal session-level pacing field
definition 7-136 description 8-9
threshold valve tpIPWT internal session-level pacing field
multifunction IOP (input/output processor) 7-26 description 8-9
Time column 7-137 tpLRUD line transmission field
time intervals description 8-10
selecting 4-4 tpLRUR receiving data field
time stamp column 11-22 description 8-10
times called column 11-16 tpNBB field
tips, tuning F-4 number of brackets started 8-7
Tns column 7-137 tpNEB field
Tns Count column 7-137 number of brackets ended 8-7
Tns/Hour column 7-137 tpNRUD line transmission field
Tns/Hour Rate column 7-137 description 8-10
TOD of Wait column 7-137 tpNRUR receiving data field
token-ring local area network database file 3-4 description 8-10
topology database update tpNSE field
control point presentation services (CPPS) 8-15 number of sessions ended 8-7
topology maintenance 8-12 tpNSS field
topology maintenance number of sessions started 8-7
APPN (advanced peer-to-peer networking) 8-12 tpQLRL transmission priority field
Display APPN Information (DSPAPPNINF) description 8-9
command 8-13 tpQNRE transmission priority field
initial topology exchange 8-13 description 8-9
node congestion updates 8-12 tpQNRL transmission priority field
obsolete topology entry removal 8-13 description 8-9
received TDUs 8-13 tpQRRR transmission priority field
topology database update (TDU) 8-12 description 8-9
transmission group (TG) update 8-12 tpSPNW session-level pacing field
Tot column 7-137 description 8-8
Tot Nbr Tns column 7-137 tpSPPW session-level pacing field
Total /Job column 7-137 description 8-8
Total characters per transaction column 7-137 tpSPST session-level pacing field
Total column 7-137 description 8-8
total CPU column 11-15, 11-16 tpSPWS session-level pacing field
description 8-8

Index X-31
tpSPWT session-level pacing field traffic priority levels
description 8-7 comparing 8-11
tpTRUD line transmission field transaction
description 8-10 boundaries 8-1, 10-4, B-1
trace analysis I/O summary 10-4 Client Access shared folders 8-33
Trace Analysis I/O Summary Report 10-6 conditions for number of 7-123
Trace Analysis Summary Report 10-5 OS/400 file server 8-33
trace data performance 10-2
analyzing 4-7 response reports B-2
collecting 7-37, 7-63 Transaction Report
collection 3-2 *DI value 8-1
creating 7-37, 7-63 *DQ value 8-1
database file 3-5 boundary values 7-39
definition 3-5 description 7-37
how to collect Job Summary Data section 7-46
DMPTRC parameter 3-10 output (QPSPDTS) 7-37
TRACE parameter 3-10 printing 7-36, 7-90, A-64
storage restrictions 3-2 RPTTYPE (*TNSACT) option 7-37, 7-45
trace event sample report 7-49, 7-62
description Transaction Response Time (Sec/Tns) column 7-138
auxiliary storage management 11-26 Transaction Significance
BASE 11-24 sample report 7-53
DASD server 11-31 section description 7-40
disk 11-26 Transactions per hour (local) column 7-138
Java 11-32 Transactions per hour (remote) column 7-139
job 11-28 Transition Report
MI complex instructions 11-34 description 7-46
page fault 11-27 output (QPSPDTD) 7-37
process 11-28 printing 7-36
program bracket 11-31 RPTTYPE (*TRSIT) option 7-37
segment address register 11-28 sample 7-46
seize lock 11-28 sample report 7-62
Trace Job (TRCJOB) command 10-4 summary 7-46
trace job information 10-4 transmission group (TG) update
Trace Job Information Report 10-7 topology maintenance 8-12
trace options transmission priority
call (external) 10-8 description 8-9
data (trace) 10-8 user-defined transmission priority 8-9
event (handler) 10-8 transmission time B-1
EXTXHINV (external exception handler) 10-8 Transmit/Receive/Average Line Util column 7-139
EXTXHRET (call termination) 10-8 TRCJOB (Trace Job) command 10-4
INTXHINV (internal exception handler) 10-8 TSE column 7-139
INTXHRET (return from an exception) 10-8 tuning
INVEXIT (call exit routine) 10-8 checklist F-1, F-2
ITERM (intervening call termination) 10-8 tips F-4
ITRMXRSG (resignaling exception) 10-8 tuning system
PTRMTPP (process termination) 10-8 advisor recommendations 4-14
PTRMUNHX (unhandled exception) 10-8 Typ column 7-139
SSPTRC (trace suspended) 10-8 Type column 7-140, 11-10, 11-11
trace periods for trace data 7-39
trace report
printing 3-5, A-66 U
trace wrap count column 11-12 UADA (user area disk activity) C-1
tracing UDR (update, delete, or release record) 10-7
job 10-4 Unit column 7-141, 11-24

X-32 Performance Tools V4R2

Unit Name column 7-141 Work with System Activity display 6-2
unknown CPU column 11-16 Work with System Status (WRKSYSSTS)
update, delete, or release record (UDR) 10-7 command 2-3
use analysis, system 3-2 Working set size (KB) column 7-143
user area disk activity (UADA) C-1 working with
user column 11-10 disk status 2-3
User ID column 7-141 functional areas A-71
User Name column 7-141 jobs 6-6
User Name/Thread column 7-141 submitted jobs 2-3
User Starts column 7-141 subsystems 2-3
User Stops column 7-141 system activity A-72
User Total column 7-141 Workload
user-based pricing 2-1 sample report 7-20
using the advisor's results 4-7 section description 7-18
Util 2 column 7-141 Writes per Second column 7-144
Util column 7-141 WRKACTJOB (Work with Active Jobs) command
utilities, process access group (PAG) 10-2 system values option 2-3
WRKDSKSTS (Work with Disk Status) command 2-3
WRKFCNARA (Work with Functional Areas)
V command A-71
valid x-axis and y-axis values 9-7 WRKJOB (Work with Job) command 2-3, 6-6
Value column 7-142 WRKPFRCOL (Work with Performance Collection)
Verify column 7-142 command 3-12, 9-16, E-1
version column 11-13 WRKSBMJOB (Work with Submitted Jobs)
command 2-3, 7-13
WRKSBS (Work with Subsystems) command 2-3
W WRKSPLF (Work with Spooled Files) command 7-14
W-I Wait/Tns column 7-142
WRKSYSACT (Work with System Activity)
Wait Code column 7-142
command A-72
wait states
description 6-1
description 7-133
WRKSYSSTS (Work with System Status)
Wait-Inel column 7-143
command 2-3
work management functions
WSC (local work station controller) 3-4
accessing 6-8
WTO column 7-144
Work Station Controller column 7-143
Work with Active Jobs (WRKACTJOB) command
system values option 2-3 X
Work with All Spooled Files display 7-14 X.25 database file 3-4
Work with Disk Status (WRKDSKSTS) command 2-3
Work with Functional Areas (WRKFCNARA)
command A-71
Work with Graph Formats and Packages display 9-3
Work with Historical Data display 9-13, 14-1
Work with Job (WRKJOB) command 2-3, 6-6
Work with Job Traces display 10-1
Work with Performance Collection (WRKPFRCOL)
command 3-12, 9-16, E-1
Work with Performance Collection display 3-12
Work with Spooled Files (WRKSPLF) command 7-14
Work with Submitted Jobs (WRKSBMJOB)
command 2-3, 7-13
Work with Subsystems (WRKSBS) command 2-3
Work with System Activity (WRKSYSACT)
command A-72
description 6-1

Index X-33
Communicating Your Comments to IBM
AS/400e series
Performance Tools for AS/400
Version 4
Publication No. SC41-5340-00

If you especially like or dislike anything about this book, please use one of the methods
listed below to send your comments to IBM. Whichever method you choose, make sure you
send your name, address, and telephone number if you would like a reply.

Feel free to comment on specific errors or omissions, accuracy, organization, subject matter,
or completeness of this book. However, the comments you send should pertain to only the
information in this manual and the way in which the information is presented. To request
additional publications, or to ask questions or make comments about the functions of IBM
products or systems, you should talk to your IBM representative or to your IBM authorized
remarketer.

When you send comments to IBM, you grant IBM a nonexclusive right to use or distribute
your comments in any way it believes appropriate without incurring any obligation to you.

If you are mailing a readers' comment form (RCF) from a country other than the United
States, you can give the RCF to the local IBM branch office or IBM representative for
postage-paid mailing.
Ÿ If you prefer to send comments by mail, use the RCF at the back of this book.
Ÿ If you prefer to send comments by FAX, use this number:
– United States and Canada: 1-800-937-3430
– Other countries: 1-507-253-5192
Ÿ If you prefer to send comments electronically, use this network ID:
– IBMMAIL(USIB56RZ)
– [email protected]

Make sure to include the following in your note:

Ÿ Title and publication number of this book
Ÿ Page number or topic to which your comment applies.
Readers' Comments — We'd Like to Hear from You
AS/400e series
Performance Tools for AS/400
Version 4
Publication No. SC41-5340-00

Overall, how satisfied are you with the information in this book?
Very Very
Satisfied Satisfied Neutral Dissatisfied Dissatisfied
Overall satisfaction Ø Ø Ø Ø Ø

How satisfied are you that the information in this book is:
Very Very
Satisfied Satisfied Neutral Dissatisfied Dissatisfied
Accurate Ø Ø Ø Ø Ø
Complete Ø Ø Ø Ø Ø
Easy to find Ø Ø Ø Ø Ø
Easy to understand Ø Ø Ø Ø Ø
Well organized Ø Ø Ø Ø Ø
Applicable to your tasks Ø Ø Ø Ø Ø

Please tell us how we can improve this book:

Thank you for your responses. May we contact you? Ø Yes Ø No

When you send comments to IBM, you grant IBM a nonexclusive right to use or distribute your comments
in any way it believes appropriate without incurring any obligation to you.

Name Address

Company or Organization

Phone No.
 Cut or Fold
Readers' Comments — We'd Like to Hear from You

IBM
Along Line
SC41-5340-00


Fold and Tape Please do not staple Fold and Tape

NO POSTAGE
NECESSARY
IF MAILED IN THE
UNITED STATES

BUSINESS REPLY MAIL

FIRST-CLASS MAIL PERMIT NO. 40 ARMONK, NEW YORK

POSTAGE WILL BE PAID BY ADDRESSEE

IBM CORPORATION
ATTN DEPT 542 IDCLERK
3605 HWY 52 N
ROCHESTER MN 55901-7829

Fold and Tape Please do not staple Fold and Tape

Cut or Fold
SC41-5340-00 Along Line
IBM 

Printed in the United States of America

on recycled paper containing 10%
recovered post-consumer fiber.

SC41-534ð-ðð
Spine information:

IBM AS/400e series Performance Tools for AS/400 Version 4