OpenSAP Frun1 Week 2 All Slides
OpenSAP Frun1 Week 2 All Slides
▪ Application purpose
▪ Application features
▪ Demo
▪ Application benefits
Monitoring
Analytics
Alerting
Availability checks
Health Monitoring provides the following TCP Network
component
availability checks to monitor the Simple
(e.g. printer)
Diagnostics
Simple
availability of network services: HTTP Cloud
Agent
Diagnostic
Agent RFC Service
ABAP
▪ HTTP Availability System
▪ RFC Availability
Firewall
▪ TCP Availability Simple HTTP Proxy
Health Diagnostics
(optional)
Web Site
Agent
The availability checks are regularly Monitoring
executed by the simple diagnostics
agents.
Simple HTTP Cloud
Diagnostics
Simple
It is possible to configure multiple simple Agent
Diagnostic
Service
Simple HTTP
diagnostics agents as a collection group. Agent
Diagnostic Web Server
This allows a load distribution and Agent
Availability checks
HTTP availability checks can be used to monitor the availability of HTTP or HTTPS URLs
Availability checks
Health Monitoring provides a CSV import function for availability metrics. This can be used for mass
configuration and update of URLs to be monitored.
Please check the following guide in the SAP Focused Run Expert Portal:
https://support.sap.com/content/dam/support/en_us/library/ssp/alm/sap-focused-run/mass-maintenance-availability-metrics-
health-monitoring_3.0_FP03.pdf
OS Monitoring
▪ Application purpose
▪ Application features
▪ Architecture of System Monitoring
▪ Template maintenance
▪ Demo
▪ Application benefits
▪ The System Monitoring application in SAP Focused Run provides an overview of the current status of technical
systems, including their associated instances, databases, and hosts.
▪ System Monitoring is based on automated checks in regular time intervals in the four categories Availability,
Performance, Exceptions, and Configuration. In each of these categories, several metrics and corresponding
thresholds can be defined per managed object.
Customer Network
Proxy
Host Agent
Simple DA
MAI Directory Self Monitoring
R
https(s) Jco RFC R
R R R R
ST-PI
R R R
Reverse
W W Streaming REST
Proxy
R R
Event Calculation
ICM
MAI_UDM_STORE Service Data Provider
Engine
R https(s) R
AS ABAP
R
AnyDB
UDM READ API SAP HANA DB
R https(s) (including HANA XS)
API for Alert R
Consumer Connector
API for Technical AS Non_ABAP
FRUN ABAP System Monitoring IS BCIA R
Instances
Introscope EM
SAP HANA DB
[email protected]
Follow all of SAP
www.sap.com/contactsap
▪ Application purpose
▪ Application features
▪ Application usage
▪ Architecture of Alert Management
▪ Outbound connection for automatic alert reactions
▪ Demo
▪ Application benefits
▪ Alert Cluster : Group of open alerts for one issue in Overview Page
landscape Alert Cluster
1 Evaluate the alert overview regarding open alerts per 2 Check the alert details regarding behavior over time, responsible
source, category, current rating, affected managed metric, and already taken actions. Check the documentation
objects, and alert duration. section and links to tools for further analysis.
Management
Alert
Notification
various channels: Email
Management
Outbound Connector
Ticket
▪ Send SNMP Traps
[email protected]
Follow all of SAP
www.sap.com/contactsap
[email protected]
Follow all of SAP
www.sap.com/contactsap
2 Add a custom
page and select the
personalization button 4 Drag the gadgets from the
“Available Views” area to the
dashboard
[email protected]
Follow all of SAP
www.sap.com/contactsap
▪ Application purpose
▪ Application features
▪ Threshold recommendation
– Metric statistics and metric forecast
– Calculation background
– Usage
▪ System anomaly prediction
– Motivation
– How does it work?
– Application usage scenario
▪ Demo
▪ Metric forecasting
▪ System anomaly prediction
Alert Detail Metric Forecast
Gain insights on which Predict future
metric causes the
predicted anomaly
© 2023 SAP SE or an SAP affiliate company. All rights reserved. ǀ PUBLIC 4
Operations Intelligence
Threshold recommendation, metric statistics, and metric forecast
Landscape C
Value
▪ Points above or below this range are outliers
Interquartile
Box Length
▪ Depending on the standard deviation of the
Range
measurements, the factor of 1.5 can be adjusted to Median
=
either tighten or loosen the chosen threshold
▪ Also the factor can be used to calculate yellow and
red thresholds
Landscape C
Objective: Predict critical anomalies in a system 30-60 minutes in advance with high confidence
Metrics showing Metrics demonstrate
baseline behavior uncharacteristic values
Predictive
Risk score
alert
Hypothesis Model
Some metrics are likely to demonstrate uncharacteristic Values of “target variable” as a function of values of other
values a few minutes before the outage. metrics that demonstrated anomalies before an outage.
© 2023 SAP SE or an SAP affiliate company. All rights reserved. ǀ PUBLIC 9
Operations Intelligence
System anomaly prediction – How does it work?
Prediction Engine
Configuration
System Anomaly
Metric Stream
Metric & Event
Receiver
Store
R-Server
System
incl. Keras and Anomaly
TensorFlow Calculate
Results Prediction Store
RCA Store
Metrics &
System Anomaly System Monitoring & Events
Prediction System Analysis
3 Detailed analysis of specific model 4 Identification of critical metrics 5 Detailed analysis of specific
versions associated with the selected within a system which contributed metrics associated along with
system to anomaly metric forecasting
[email protected]
Follow all of SAP
www.sap.com/contactsap