Fos Ficon Admingd v730
Fos Ficon Admingd v730
15 August 2014
FICON
Administrator's Guide
Brocade, the B-wing symbol, Brocade Assurance, ADX, AnyIO, DCX, Fabric OS, FastIron, HyperEdge, ICX, MLX, MyBrocade, NetIron,
OpenScript, VCS, VDX, and Vyatta are registered trademarks, and The Effortless Network and the On-Demand Data Center are trademarks
of Brocade Communications Systems, Inc., in the United States and in other countries. Other brands and product names mentioned may be
trademarks of others.
Notice: This document is for informational purposes only and does not set forth any warranty, expressed or implied, concerning any
equipment, equipment feature, or service offered or to be offered by Brocade. Brocade reserves the right to make changes to this document
at any time, without notice, and assumes no responsibility for its use. This informational document describes features that may not be
currently available. Contact a Brocade sales office for information on feature and product availability. Export of technical data contained in
this document may require an export license from the United States government.
The authors and Brocade Communications Systems, Inc. assume no liability or responsibility to any person or entity with respect to the
accuracy of this document or any loss, cost, liability, or damages arising from the information contained herein or the computer programs that
accompany it.
The product described by this document may contain open source software covered by the GNU General Public License or other open
source license agreements. To find out which open source software is included in Brocade products, view the licensing terms applicable to
the open source software, and obtain a copy of the programming source code, please visit http://www.brocade.com/support/oscd.
Contents
Preface..................................................................................................................................... 7
Document conventions......................................................................................7
Text formatting conventions.................................................................. 7
Command syntax conventions.............................................................. 7
Notes, cautions, and warnings.............................................................. 8
Brocade resources............................................................................................ 9
Contacting Brocade Technical Support.............................................................9
Document feedback........................................................................................ 10
Introducing FICON...................................................................................................................13
FICON overview..............................................................................................13
Fabric OS support for FICON..............................................................13
Latency guideline................................................................................ 14
FICON concepts..............................................................................................15
FICON configurations......................................................................................16
Switched point-to-point........................................................................16
Cascaded FICON................................................................................17
Access control in FICON.................................................................................25
Cascaded zoning................................................................................ 26
Error reporting..................................................................................... 27
Secure access control.........................................................................28
FICON commands.......................................................................................... 30
Link and FC addressing.................................................................................. 32
Domain ID........................................................................................... 33
Port area............................................................................................. 33
ALPA................................................................................................... 33
Index.................................................................................................................................. 123
● Document conventions......................................................................................................7
● Brocade resources............................................................................................................ 9
● Contacting Brocade Technical Support.............................................................................9
● Document feedback........................................................................................................ 10
Document conventions
The document conventions describe text formatting conventions, command syntax conventions, and
important notice formats used in Brocade technical documentation.
Format Description
bold text Identifies command names
Identifies keywords and operands
Identifies the names of user-manipulated GUI elements
Identifies text to enter at the GUI
Convention Description
bold text Identifies command names, keywords, and command options.
italic text Identifies a variable.
Convention Description
value In Fibre Channel products, a fixed value provided as input to a command
option is printed in plain text, for example, --show WWN.
NOTE
A Note provides a tip, guidance, or advice, emphasizes important information, or provides a reference
to related information.
ATTENTION
An Attention statement indicates a stronger note, for example, to alert you when traffic might be
interrupted or the device might reboot.
CAUTION
A Caution statement alerts you to situations that can be potentially hazardous to you or cause
damage to hardware, firmware, software, or data.
DANGER
A Danger statement indicates conditions or situations that can be potentially lethal or
extremely hazardous to you. Safety labels are also attached directly to products to warn of
these conditions or situations.
Brocade resources
Visit the Brocade website to locate related documentation for your product and additional Brocade
resources.
You can download additional publications supporting your product at www.brocade.com. Select the
Brocade Products tab to locate your product, then click the Brocade product name or image to open the
individual product page. The user manuals are available in the resources module at the bottom of the
page under the Documentation category.
To get up-to-the-minute information on Brocade products and resources, go to MyBrocade. You can
register at no cost to obtain a user ID and password.
Release notes are available on MyBrocade under Product Downloads.
White papers, online demonstrations, and data sheets are available through the Brocade website.
Brocade customers
For product support information and the latest information on contacting the Technical Assistance
Center, go to http://www.brocade.com/services-support/index.html.
If you have purchased Brocade product support directly from Brocade, use one of the following methods
to contact the Brocade Technical Assistance Center 24x7.
Preferred method of contact for non- Required for Sev 1-Critical and Sev support@brocade.com
urgent issues: 2-High issues:
Please include:
• My Cases through MyBrocade • Continental US: 1-800-752-8061
• Problem summary
• Software downloads and licensing • Europe, Middle East, Africa, and
• Serial number
tools Asia Pacific: +800-AT FIBREE
(+800 28 34 27 33) • Installation details
• Knowledge Base
• For areas unable to access toll • Environment description
free number: +1-408-333-6061
• Toll-free numbers are available in
many countries.
• Brocade Supplemental Support augments your existing OEM support contract, providing direct
access to Brocade expertise. For more information, contact Brocade or your OEM.
• For questions regarding service levels and response times, contact your OEM/Solution Provider.
Document feedback
To send feedback and report errors in the documentation you can use the feedback form posted with
the document or you can e-mail the documentation team.
Quality is our first concern at Brocade and we have made every effort to ensure the accuracy and
completeness of this document. However, if you find an error or an omission, or you think that a topic
needs further development, we want to hear from you. You can provide feedback in two ways:
• Through the online feedback form in the HTML documents posted on www.brocade.com.
• By sending your feedback to documentation@brocade.com.
Provide the publication title, part number, and as much detail as possible, including the topic heading
and page number if applicable, as well as your suggestions for improvement.
• Added notes wherever configupload command mentioned that it backs up Fabric OS feature and
switch configuration, but not FMS and FICON-specific configuration.
• Added notes wherever enabling FMS mode is described that all high-integrity fabric attributes must
be configured and HIF mode enabled to enable FMS mode.
• Changed Moving ports to a logical switch section to Disabling and enabling ports - persistent states
on page 48 and added details comparing port persistent enable and disable states with
interactions of ASM Mode setting, IPL state, and use of portdisable and portenable commands.
• Added note in Upgrade considerations on page 61 that upgrading the Fabric OS v7.3.0 and later
from v7.2.1 and earlier requires high-integrity fabric attributes configured and HIF mode enabled.
• Added note under "Routing policies" bullet in Fabric OS support for FICON on page 13 that
System z FICON does not support exchange-based routing.
• Added note under "Insistent Domain ID (IDID)" bullet that IDID is the recommended best practice
for single-byte addressing.
• Added Extended fabric configurations on page 19.
• Added sentence to Firmware download disruption on page 79 about FCIP Hot Code Load feature
on Brocade 7840 switch.
• Added Monitoring and Alerting Policy Suite on page 89.
• In Basic Switch Configuration, changed step to enable port-based routing policy to device-based
routing policy.
● FICON overview..............................................................................................................13
● FICON concepts..............................................................................................................15
● FICON configurations......................................................................................................16
● Access control in FICON.................................................................................................25
● FICON commands.......................................................................................................... 30
● Link and FC addressing.................................................................................................. 32
FICON overview
IBM Fibre Connection (FICON®) is an industry-standard, high-speed input/output (I/O) interface for
mainframe connections to storage devices. This guide discusses support offered by Fabric OS in
intermix mode operations, in which FICON and Fibre Channel technology work together.
For specific information about intermix mode and other aspects of FICON, refer to the IBM Redbook,
FICON® Implementation Guide (SG24-6497-01), and Implementing an IBM/Brocade SAN with 8 Gbps
Directors and Switches (SG24-6116-08).
NOTE
In this guide, the term switch is used to refer to a Brocade switch, Backbone, or backbone platform
unless otherwise noted.
Disables the dynamic domain ID feature and only allows the switch to use a pre-set domain ID. All
switches in a fabric must have a unique domain ID. An insistent domain ID is required with 2-byte
addressing. IDID is the recommended best practice for single byte addressing.
• Link incident detection, registration, and reporting
Provides administrative and diagnostic information.
• Swap port area IDs (PIDs) of physical ports
Redirects resources from a failed port to a healthy port without changing the mainframe hardware
configuration definition (HCD) settings. This feature, also called "port swapping," is available for
both FICON and open system environments. Swapping PIDs on ports resolves situations in which
the hardware has failed and the channel configurations cannot be changed quickly. Port swpping
has minimal or no impact on other switch features.
• Switch connection control (SCC) policy
Includes fabric security methods that prevent unauthorized switches from joining a fabric. SCC
policy is required for cascaded FICON configurations and whenever 2-byte addressing is used.
• Traffic Isolation (TI) Zones and Enhanced TI Zones
TI Zones are used to direct traffic across links through a specified path. Enhanced TI Zones allow
you to have ports in more than one TI Zone and to program domain controller routes to destination
domains for F-class traffic, ensuring fabric stability.
NOTE
For more detail on these features and configuration procedures, refer to the Fabric OS Administrator's
Guide.
Brocade management tools provide further support:
• Brocade Network Advisor
Brocade Network Advisor is an optional software program that can be used to manage a fabric that
supports FICON and Fibre Channel Protocol (FCP) devices and traffic. This is the recommended
GUI management tool for FICON environments on B-series enterprise-class switches. For more
information on Brocade Network Advisor, refer to the manual appropriate for your version
requirements:
‐ Brocade Network Advisor SAN + IP User Manual
‐ Brocade Network Advisor SAN User Manual
• Web Tools
Web Tools is an embedded GUI-management tool that can be used to manage a Brocade switch or
backbone that supports FICON and Fibre Channel Protocol (FCP) devices and traffic. For more
information on Web Tools, refer to the Web Tools Administrator's Guide.
Latency guideline
The maximum supported distance for a FICON channel is 300 Km (1.5 msec of delay). Synchronous
mirroring applications are generally limited to 100 Km (0.5 msec of delay). Greater distances require
that the FICON Acceleration feature be used with FCIP. The FICON Acceleration feature emulates
control unit response to the channel to make the devices appear closer to the channel than they
actually are.
FICON concepts
The following figure shows how the traffic in a switched point-to-point configuration flows in a FICON
environment. The logical path of the traffic is defined as frames moving from the channel to the switch
to the control unit. FICON traffic moves from a logical partition (LPAR) and through the channel, through
a Fibre Channel link to the switch through the control unit, and ends at the device. This is also called a
channel path , which is a single interface between a central processor and one or more control units
along which signals and data can be sent to perform I/O requests. The channel path uses the logical
path to traverse the Fibre Channel fabric. The channel path is defined using an ID, called the channel
path ID (CHPID). This information is stored in the Input/Output Definition File (IODF) and may be
dynamically configured using the mainframe feature, zDAC (Discover and Auto Configuration). The
IODF is typically built using the hardware configuration definition (HCD).
The traffic on the channel path communicates using channel command words (CCWs) that direct the
device to perform device-specific actions, such as seek, read, or rewind. In a FICON environment,
buffer credits are used at the Fibre Channel Protocol (FCP) level for flow control between optically
adjacent ports, while information unit (IU) pacing is the flow control mechanism used by the channel.
There are times when there are no more buffer credits to pass back to the other end of the link and a
frame pacing delay occurs. Frame pacing delay is the number of intervals of 2.5 microseconds that a
frame had to wait to be transmitted due to a lack of available buffer credits. Frame pacing delay
information is reported in the FICON Director Activity Report with the System z RMF feature.
FICON introduces the following concepts:
• FICON Control Unit Port (CUP)
The internal port in a switch that assumes an Fibre Channel (FC) address such that it is the FC
domain ID (DID) used to direct FICON traffic to the FICON Management Server (FMS).
• FICON Manager
Host communication includes control functions such as blocking and unblocking ports, as well as
monitoring and error-reporting functions.
• Hardware configuration definition (HCD)
HCD is an IBM interactive interface application that allows you to define the hardware configuration
for both the processor channel subsystem and the operating system running on the processor.
• Information unit
A unit of FICON data consisting of from one to four Fibre Channel frames.
• Link Incident Record Registration (LIRR)
The LIRR Extended Link Service (ELS) requests that the recipient add the requesting port to its list
of ports that are to receive a Registered Link Incident Report (RLIR).
• Node
A node is an endpoint that contains information. It can be a computer (host), a device controller, or
a peripheral device, such as a disk array or tape drive. A node has a unique 64-bit identifier known
as the Node_Name. The Node_Name is typically used for management purposes.
• Prohibit Dynamic Connectivity Mask (PDCM) and connectivity attributes
PDCM controls whether communication between a pair of ports in the switch is prohibited.
Connectivity attributes control whether all the communication is blocked for a port.
• Read Record Set (RRS)
RRS is an IBM Channel-initiated CCW command. The Brocade FCIP FICON Acceleration License
allows the emulation of command chains that include this CCW command. The command is used in
IBM z/OS Global Mirror configurations to read updates from a volume in an active mirroring session.
• Registered Link Incident Report (RLIR)
RLIR ELS provides a way for a node port to send an incident record to another node port.
• Request Node Identification Data (RNID)
RNID ELS acquires the associated node’s identification data, which provides configuration
discovery and management purpose information.
• Resource Measurement Facility (RMF)
Performance monitoring software that gathers transaction data from the environment and generates
performance reports. All Level II reports, which include port statistics, require the FICON Control
Unit Port (CUP) and FICON Management Server (FMS).
• Systems Operations (SysOps)
SysOps provides the ability to monitor and control all subsystems in a sysplex from any system in
the sysplex. This includes controlled startup, controlled shutdown, and automated recovery of
software resources.
• Sysplex
In IBM mainframe computers, a Systems Complex, commonly called a sysplex, allows multiple
processors to be joined into a single unit, sharing the same sysplex name and Couple Data Sets.
• IPL file
The initial program load (IPL) File, located in nonvolatile storage, contains the current, active
configuration settings for the FICON director. If functions to initialize data on the Director during a
POR event. When the "Active=Saved" FICON CUP mode register bit setting is on, any active
configuration in switch memory is automatically saved to the IPL file.
FICON configurations
There are two types of FICON configurations that are supported using Brocade Fabric OS: switched
point-to-point and cascaded topologies.
Switched point-to-point
A single-switch configuration is called switched point-to-point allows the channel to use single-byte
addressing.
Cascaded FICON
Cascaded FICON refers to an implementation of FICON that uses one or more FICON channel paths in
which the domain ID of the entry switch is different than the domain ID of the switch where the control
unit is attached. Therefore, cascading requires a two-byte link address. Anytime a two-byte link address
is defined on a channel, all link addresses must be two-byte link addresses.
Switches may be interconnected using the following links:
• Traditional Inter-Switch Links (ISLs)
• Inter-Chassis Links (ICLs)
• Fibre Channel over Internet Protocol (FCIP)
The processor interface is connected to one switch (known as the entry switch), while the storage
interface is connected to the other. This configuration is supported for both disk and tape, with multiple
processors, disk subsystems, and tape subsystems sharing the ISLs or ICLs between the switches and
backbones. Multiple ISLs between the switches and backbones are also supported. Cascading between
switches and backbones is also supported. An example of this would be a Brocade DCX 8510-8
Backbone connected to a Brocade 6510.
A cascaded configuration (refer to the following figure) requires two-byte addressing. Two byte-
addressing requires a list of authorized switches. This authorization feature, called fabric binding, is
available through the Secure Access Control List feature. The fabric binding policy allows a predefined
list of switches (domains) to exist in the fabric and prevents other switches from joining the fabric. This
type of configuration is described in User security considerations on page 35.
The following figures show two cascaded configurations. These configurations require Channel A to be
configured for two-byte addressing and require IDID and fabric binding. It is recommended that there be
only two domains in a path from a FICON Channel interface to a FICON Control Unit interface. There
are exceptions to the two-domain rule in extended fabric configurations. Refer to Extended fabric
configurations on page 19 for examples.
The following figure illustrates multiple switches cascaded off of switch 21. As long as there is only one
hop from channel to control unit, the configuration is supported.
Switches in cascaded configurations may be connected through interchassis links (ICLs), interswitch
links (ICLs), and FCIP. Connection using FCIP is through Fibre Channel extension devices, such as
7800 switches and FX8-24 blades. Following are example configurations.
For more information on long distance and extended fabrics, refer to the Fabric OS Administrator's
Guide. For more information on FCIP and extension products, refer to the Fabric OS FCIP
Administrator's Guide.
The following figure illustrates a multi-hop ICL triangle configuration that uses ICLs. Note that three
switches are connected through ICLs only.
The following figure illustrates a multi-hop configuration that uses ICLs and ISLs. This configuration is
supported with or without switches 4 or 1. All switches must be all generation (Gen) 4 or all Gen 5.
You cannot mix Gen 4 and Gen 4 on ICL connections.
The following figure illustrates a multi-hop configuration that uses ICLs and FCIP. The two 7800
switches are for ISL extension only using FCIP. Channel or control unit connections are not permitted.
These switches may be replaced with an FX8-24 blade installed in directors. This configuration can be
supported with or without switches 4 or 1. All switches must be Gen 4 or Gen 5. You cannot mix Gen 4
and Gen 5 switches with ICLs.
The following figure illustrates a configuration that uses FCIP between 7800 switches that are used as
routers only, for ISL extension. Channel or control unit connections are not permitted. The 7800
switches may be replaced with FX8-24 blades installed in directors.
The following figure illustrates a configuration that uses cascaded FCIP with four 7800 switches. The
7800 switches at site 1 are used for ISL extension only. Channel or control unit connections are not
permitted. The 7800 switches at site 2 are only for network, channel, and control unit connections. The
7800 switches on either site may be replaced with the FX8-24 blades installed in directors.
The following figure illustrates a cascaded configuration that uses FCIP and 7800 switches as edge
switches. At site 1, the 7800 switches are used for ISL extension only. Channel or control unit
connections are not permitted. At site 2, only channel, and control unit connections are permitted on
the 7800 switches. The 7800 switches at either site may be replaced with the FX8-24 blades installed
in directors.
NOTE
Session-based zoning enforcement is not recommended on a FICON switch. For more information on
session-based zoning enforcement, refer to the Fabric OS Administrator's Guide.
Cascaded zoning
The figure below illustrates multiple sites sharing the same disaster-recovery site. Each switch or
Backbone at a remote site, labeled Director 1 and Director 31, can pass traffic to Director 2, but no
traffic is permitted between Zone A and Zone B.
The figure below illustrates the multiple zoning concepts that can be used to restrict traffic. Any host
channel at the Backup Site (connected to Director 11 or Director 12) can connect to the backup tape
contained within the same zone. Notice that no more than a single hop is ever allowed and only
Channel Path Identifiers (CHPIDs) 79 and 7A on the Primary Site can connect to the backup tape.
Furthermore, CHPIDs 79 and 7A can only connect to the backup tape at the Backup Site.
NOTE
Zoning does not replace the need to set up the connectivity from the host to storage control units in
the HCD or Input/Output Configuration Program (IOCP). For more information on zoning, refer to the
Fabric OS Administrator's Guide.
Zone A (Blue): Any CHPID connected to Director 1, except CHPID 79, can get to any control unit
connected to Director 1. The zone includes all ports in Director 1 except ports 4, 5, and 6.
Zone B (Orange): Any CHPID connected to Director 2, except CHPID 7A, can get to any control unit
connected to Director 2. The zone includes all ports in Director 2 except ports 4, 5, and 6.
Zone C (Green): Any CHPID connected to Director 11 can get to any control unit connected to Director
11. The zone includes all ports in Director 11 except ports 5 and 6. Adding ports 5 and 6 to the zone, so
that all ports in the switch or Backbone are in the same zone, would not affect permitted connectivity
and may be a more practical alternative.
Zone D (Yellow): Any CHPID connected to Director 12 can get to any control unit connected to Director
12. The zone includes all ports in Director 12 except ports 5 and 6, which are used for ISLs.
Red Zone E: CHPID 79 can talk only to the remote tape connected to ports 7 and 8 on Director 11. The
zone includes port 4 of Director 1 and ports 7 and 8 of Director 11. Either ISL can be used.
Purple Zone F: CHPID 7A can talk only to the remote tape connected to ports 7 and 8 on Director 12.
The zone includes port 4 of Director 2 and ports 7 and 8 of Director 12. Either ISL can be used.
Error reporting
Non-implicit link incidents (such as Fabric OS recognized or bit error rate threshold exceeded) and
implicit link incidents (such as FRU failure) are reported to registered listeners on the local switch. The
RMF 74-7 record (FICON Director Activity Report, which is the same RMF record containing the
average frame pacing delay information) reports port errors, which in turn are also reported back to the
mainframe host management consoles.
FICON commands
NOTE
The Fabric OS CLI supports only a subset of the Brocade management features for FICON fabrics.
The full set of FICON CUP administrative procedures is available using the Brocade Network Advisor
and Web Tools software features. You can also use an SNMP agent and the FICON Management
Information Base (MIB).
The table below summarizes the Fabric OS CLI commands that can be used for managing FICON
fabrics. For detailed information on these commands, refer to the Fabric OS Command Reference.
Command Description
Command Description
ficonClear rlir Removes all RLIR records from the local RLIR
database.
ficonClear rnid Removes all outdated RNID records from the local
RNID database.
ficonCfg --set LIRR portnumber Sets the current LIRR device port number persistently.
ficonCfg --reset LIRR portnumber Clears the currently configured LIRR port number.
ficonCupSet fmsmode enable | disable |reset Enables, disables, or resets FICON Management
Server (FMS) mode.
CAUTION
Use the ficonCupSet fmsmode reset option only
under direction from technical support personnel.
Indiscriminate use can disrupt CCW processing
and create errors at the host.
ficonCupSet CRP PID CHID Sets the current reporting path (CRP).
ficonCupShow DD_LOG Displays the log information associated with the last
diagnostic command processed by the CUP.
Command Description
ficonCupShow hlthchk_log Displays health check logs for the logical switch.
ficonShow lirr fabric Displays registered listeners for link incidents for the
local switch or for the fabric, if specified.
ficonShow rnid fabric Displays node identification data for all devices
registered with the local switch or all devices
registered with all switches defined in the fabric, if
specified.
ficonShow rnid port Displays node identification data for a specified port.
ficonShow switchrnid fabric Displays node identification data for the local switch or
for the fabric, if specified.
For limitations and considerations for using Fabric OS commands with FMS mode enabled, refer to
Fabric OS command limitations and considerations on page 59.
Domain ID
Although you enter the domain ID in decimal format when configuring a switch, it is represented in
hexadecimal formal in the FC address.
For single-byte addressing, the domain area returned from the switch where the channel logs in is used
for the FC address. Therefore, the channel and control unit must be in the same logical switch. With
two-byte link addressing, the most significant byte of the link address is used for the domain area of the
FC address. Once two-byte link addressing is defined for a channel, all link addressing for that channel
must use two-byte link addressing.
Port area
The port area address is the single-byte link address or the least significant byte of a two-byte link
address. The link address is entered in the HCD in hexadecimal format and is represented in
hexadecimal in the FC address.
ALPA
The Arbitrated Loop Physical Address (ALPA) was originally used in fibre channel for loop devices.
Currently, the ALPA is used for N_Port ID Virtualization (NPIV), which allows multiple WWNs to log in to
the same switch port. The ALPA determines the logical entity to which frames belong. This is how
virtual servers using zLinux or zVM can share the same channel. Because the DCX and DCX 8510-8
Backbones can have up to 512 ports, the upper two bits of the ALPA are used in certain addressing
modes.
FICON does not use the ALPA byte. However, the ALPA is a required byte in the FC address. The
channel completes the FC address for a control unit link address by acquiring the ALPA that the switch
returned to the channel when the channel logged in. This is why the ALPA must be the same for both
the channel port and the control unit port. This is also the reason why 10-bit addressing mode cannot be
used for FICON (refer to Addressing modes on page 46).
NOTE
You enable the fabric-wide consistency policy on the fabric once the switch joins the fabric.
NOTE
If FMS mode is enabled before upgrading to v7.3.0, IDID, SCC_Policy, and SCC:S will be validated
and the firmware attempt failed if either are incorrect. If validation is successful, HIF mode will
automatically enable when the firmware installs.
If a FICON channel tries to connect to a fabric switch without these features configured, the channel
segments from the fabric.
Once these features are configured, you must enable the switch in High-Integrity Fabric (HIF) mode
using the Fabric OS configure command. This verifies the required features are set and locks the
configuration to ensure connection with the FICON channel. Once the HIF mode is set, you cannot
change the IDID, fabric-wide consistency policy, and SCC policy without disabling HIF mode.
Following are considerations for using HIF mode:
• You must enable HIF mode to enable FMS mode.
• Before a Fabric OS downgrade, you must disable HIF mode. Note that this operation is not
recommended for FICON and should only be used when downgrading firmware. You will receive a
warning to this effect if FMS mode is enabled. If HIF is disabled, any new channel initialization
would fail as the Query Security Attribute (QSA) reply from the switch to the channel will fail. The
existing channel will continue to operate, however.
• Before a Fabric OS upgrade, be sure the switch has appropriate IDID, fabric-wide consistency
policy, SCC policy settings are enabled so that HIF mode can enable when the firmware installs.
The following instructions are provided in this section to configure a switch as part of a high-integrity
fabric:
• Enabling insistent domain ID on page 36
• Creating and activating the SCC policy on page 36
• Enabling the fabric-wide consistency policy on page 37
• Enabling High-Integrity Fabric mode on page 37
Creating a switch connection control (SCC) policy defines switches allowed in the fabric.
To configure and activate an SCC policy, use the following steps.
1. Connect to the switch and log in.
2. Perform one of the following steps:
• Enter the secpolicycreate command to add all switches in the fabric, if they are connected.
secpolicycreate "SCC_POLICY","*"
• Enter the secpolicyadd command to add one or more members to an existing policy. The
following command is an example of adding a member using device WWNs.
secpolicyadd "SCC_POLICY","wwn1;wwn2"
3. Enter the secpolicyactivate command to activate the currently defined SCC policy.
This activates the policy set on the local switch or all switches in the fabric, depending on the
configured fabric-wide consistency policy.
For a list of high-integrity fabric requirements for switches, refer to Meeting high-integrity fabric
requirements on page 35. Setting HIF mode locks the IDID, fabric-wide consistency policy, and SCC
policy settings to ensure that the fabric is of high integrity so that it can connect with the FICON
channel. You cannot change these settings without disabling HIF mode.
NOTE
HIF mode must be enabled to enable FMS mode.
If HIF configuration requirements have not been met, an error message describes what you must
configure for the command to succeed. For example, the following message states that an IDID, SCC
policy or fabric-wide consistency policy have not been configured for the switch. Perform additional
configuration if required, then enable HIF mode.
Error: Unable to set HIF Mode. No valid IDID settings,
SCC policy and/or Fabric wide(SCC:S) configuration
secpolicyactivate
portenable n
NOTE
For more detailed information on commands and command output referenced in this section, refer to
the Fabric OS Command Reference.
CAUTION
Configuring the switch for FICON is a disruptive process. The switch must be disabled to
configure switch parameters.
1. Connect to the switch and log in using an account assigned to the admin role.
2. Enter the switchDisable command. You will need to disable the switch to access all the switch
parameters.
3. Configure the switch and chassis name, if applicable.
4. Set the routing policy by entering the aptPolicy command.
Port-based routing (PBR) and device-based routing (DBR) are qualified for System z; however,
reference your system qualification letter for current support information. The recommended policy is
DBR (aptPolicy 2).
If FICON Emulation features (IBM z/OS Global Mirror or Tape Pipelining) are enabled on an FCIP
Tunnel in the switch, PBR or (aptPolicy 1) must be used.
5. Configure Dynamic Load Sharing.
The recommended best practice is to enable Dynamic Load Sharing (DLS); however, DLS is only
supported when Lossless is enabled.
• To enable Lossless with DLS, use the dlsSet --enable -lossless command.
• If Lossless will not be used, use the dlsReset command.
NOTE
If Lossless DLS is not enabled, the routing policy must be port-based routing (aptPolicy 1).
6. Set In-Order Delivery using the iodSet command.
7. Configure the switch parameters using the configure command and enter the responses shown in
the table below when prompted. (Items in italics are top-level parameters.)
TABLE 2 FICON switch parameters
Domain The domain ID is the switch address. The recommended best practice is to
set the switch address and domain ID to be the same. The domain ID is
entered in decimal, so a switch address of hexadecimal 0x22 would have a
domain ID of decimal 34.
R_A_TOV 10000 Do not change. The Resource Allocation TimeOut Value (R_A_TOV) is
entered in milliseconds. This works with the variable E_D_TOV to determine
switch actions when presented with an error condition.
Allocated circuit resources with detected errors are not released until the
time value has expired. If the condition is resolved prior to the timeout, the
internal timeout clock resets and waits for the next error condition.
E_D_TOV 2000 The Error Detect TimeOut Value (E_D_TOV) is a timeout value entered in
milliseconds. This timer is used to flag a potential error condition when an
expected response is not received within the set time limit. If the time for an
expected response exceeds the set value, then an error condition occurs.
Leave this value at 2 seconds (2000) unless connected to extension
equipment. In some cases, when connecting to extension equipment, such
as the Brocade 7800, it must be set to 5 seconds (5000). This field should
only be changed after consulting a Brocade Certified Fabric Professional.
MAX HOPS 7 This parameter should not be confused with the one- hop FICON cascading
limitation. There are configurations that are more than one hop because
more than two domain IDs are used in the path that fit within the IBM
support limitation of "one hop."
Data field size 2112 Do not change. Specifies the largest possible value, in bytes, for the size of
a type 1 (data) frame. The switch advertises this value to other switches in
the fabric during construction of the fabric and to other devices when they
connect to the fabric. Setting this value smaller than 2112 may result in
decreased performance.
Sequence level 0 Do not change. When set to 1, frames of the same sequence from a
switching particular source are transmitted together as a group. When set to 0, frames
are interleaved among multiple sequences when transmitted. Under normal
conditions, sequence-level switching should be disabled for better
performance.
Disable device 1 The recommended best practice is to disable device probing. When disable
probing device probing is set to 1, devices that do not register with the Name Server
will not be present in the Name Server database.
Suppress Class F 0 Do not change. By default, the switch can send Class F frames. When this
traffic option is turned on (set to 1), Class F traffic is converted to Class 2 traffic
before being transmitted to support remote fabrics that involve ATM
gateways.
Per-frame route 0 Do not change. In addition to the virtual channels used in frame routing
priority priority, when this value is set to 1, support is also available for per-frame-
based prioritization.
Insistent Domain ID Yes The recommended best practice is to set the domain ID to be insistent.
Mode Setting the insistent domain ID is required for two-byte addressing.
High Integrity No Before setting HIF mode, the following attributes must be configured:
Fabric Mode
• An insistent domain ID (IDID
• A Fabric Wide Consistency Policy => SCC:S (Strict mode)
• A valid SCC_Policy (configured and activated)
Setting HIF mode locks the IDID, fabric wide consistency policy, and SCC
policy settings to ensure that the fabric is of high integrity so that it can
connect with the FICON channel. You cannot change these settings without
disabling HIF mode.
Virtual Channel No Do not change. The switch allows fine-tuning of a specific application
parameters through virtual channel parameters. The default virtual channel settings
have already been optimized for switch performance. Changing these
values can improve switch performance, but can also degrade performance.
F-Port login No With Fabric OS v6.1.0 and later, it is not necessary to make any changes to
parameters F_Port login parameters. The only other reason to answer yes (y) to this
prompt is if NPIV is being used and there is a need to change the default
number of NPIV logins.
Maximum logins Change this only if NPIV is being used and the number of fabric logins will
per switch exceed the default or there is a need to limit the number of logins.
Maximum logins Change this only if NPIV is being used and the number of port logins will
per port exceed the default or there is a need to limit the number of logins.
Maximum logins For Fabric OS v6.1.0b and later - Do not change. Leave at default.
per second
Login stage interval Do not change. Specifies the stage interval in the staged F_Port login, in
milliseconds.
Zoning operation No Do not change. Disables and enables NodeName Zone Checking.
parameters
RSCN transmission No Configures end-device RSCN Transmission Mode values such as:
mode
• RSCN only contains single PID
• RSCN contains multiple PIDs (Default)
• Fabric addresses RSCN
Arbitrated Loop No Parameters include Send FAN frames, Enable CLOSE on OPEN received,
parameters and Always send RSCN. Refer to the configure command in the Fabric OS
Command Reference for more information.
System services No Enables or disables the read link status (RLS) probing performed by the
FCP daemon.
Portlog events No Determines whether various types of port events are logged.
enable
RPCD attributes No Enables or disables Remote Procedure Call Daemon (RPCD) attributes.
cfgload attributes No Configures configuration upload and download parameters. Refer to the
configureChassis command in the Fabric OS Command Reference for
more information.
Web Tools No Modifies Webtools attributes such as enabling the Basic User Enabled
attributes Perform License, Checking and Warning, Allow Fabric Event Collection, and
Login Session Timeout.
NOTE
8 Gbps ports will not auto-negotiate to 1 Gbps and cannot be configured for 1 Gbps. 16 Gbps ports
will not auto-negotiate to 2 Gbps or 1 Gbps.
12.Enter the portCfgLongDistance command for ports with fiber connections exceeding 10 km.
For ports with fiber connections exceeding 10 km, configuring the port for long distance mode
increases the number of buffer-to-buffer (BB) credits available on that port. If you have any
Extended Fabrics links, enable VC translation link initialization to stabilize them. Refer to the Fabric
OS Administrator's Guide for details on this option of the portCfgLongDistance command.
13.Configure the fill word on Gen4 (8 Gbps capable) switches using the portCfgFillWord command.
The following example shows the command with mode 3 and passive operands set:
portcfgfillword slot/port, 3 passive
NOTE
This command is only supported on Gen4 (8 Gbps capable) switch products. It is not supported on
Gen2 (4 Gbps capable) switches and is automatic in Gen5 (16 Gbps capable) switches. Refer to your
switch vendor for appropriate fill word settings, such as mode. For more information on the
portCfgFillWord command, refer to the Fabric OS Command Reference.
14.Enter the defZone --noaccess command to disable the default zone.
15.Configure zoning for your fabric.
Follow standard FICON zoning practices published by IBM. For more information on zoning, refer to
the administrator’s guide for your management tool.
16.Enter the statsClear command to clear port statistics.
17.Enter the switchShow command to verify that the switch and devices are online.
18.Enter the ficonShow rnid command to verify that the FICON devices are registered with the switch.
Enter this command only if fabric binding is enabled; otherwise, channels will be attached as invalid.
The ficonShow rnid port command displays RNID information for a port.
19.Enter the ficonShow lirr command to verify that the FICON host channels are registered to listen for
link incidents. Enter this command only if fabric binding is enabled; otherwise, channels will be
attached as invalid.
20.Enter the configUpload command to save your configuration.
NOTE
This command backs up the current FOS feature and switch configuration, but does not back up
FMS or FICON-specific configuration such as IPL file, active configuration, mode register, and host
data.
21.Enter the command appropriate for your environment:
• If you do not have access to an FTP server, use a Telnet session that can log console output and
enter the supportShow command to save a FICON baseline.
• If you have access to an FTP server, enter the supportSave command to save the FICON
baseline of your switch.
22.Take the appropriate action based on your configuration:
• If you have a cascaded configuration, continue configuring your switches using the instructions in
Configuring cascaded FICON on page 43.
• If you have and plan to use a CUP license, continue configuring your switches using the
instructions in Configuring FICON CUP on page 57.
• If you have a switched point-to-point configuration, you have completed the switch configuration.
CAUTION
Configuring the switch for FICON is a disruptive process. The switch must be disabled to
configure switch parameters. This procedure must be followed prior to channels joining the
fabric.
In the command syntax, member indicates a switch that is permitted to join the fabric. Specify
switches by WWN, domain ID, or switch name. Enter an asterisk (*) to indicate all the switches in
the fabric. To create a policy that includes all the switches in the fabric, enter the following
command:
For Fabric OS v7.3 and later, FICON cascading is allowed in an FCR configuration when front or
translate domains are present. These domains can co-exit with a fabric-wide "strict" SCC policy.
To connect edge fabrics to high-integrity backbone fabrics that use a FCR, use the following these
general steps:
1. Define the FC Router EX_Ports.
2. Enter the portcfgexport slot/port command to obtain the front domain WWN.
3. Add the front domain WWN to the SCC Policy lists in the edge switches to be merged.
4. Connect the switches.
Switch or backbone Maximum number of logical switches Maximum number of logical switches with
FMS (CUP) enabled
DCX 8510-8 8 4
DCX 8 4
DCX-4S 8 4
5300 4 2
link feature. It is shared by multiple logical switches and can be used for multiple Fabric IDs (FIDs).
Creating an XISL is disruptive because it provides a core routing function for the switch. For more
information on creating XISLs and restrictions, refer to the configure command in the Fabric OS
Command Reference.
When using XISLs, the effective routing policy for all logical switches configured to allow XISL use
is that of the base switch regardless of how the routing policy is defined. The best practice is to set
the routing policy in all logical switches that allow XISL use to have the same routing policy as the
base switch to avoid confusion about the routing policy used. A base switch can only be used for
XISLs. Therefore, the switch cannot be managed through a CUP port and cannot report statistics
for RMF.
Port-based routing (PBR) and device-based routing (DBR) are qualified for System z; however,
reference your system qualification letter for current support information. The recommended policy
is DBR or aptPolicy 2.
• If addresses 0xFE or 0XFF have been assigned to physical ports, those ports must be disabled
before FMS can be enabled.
• Each logical switch has its own configuration data, such as the following:
‐ Initial Program Load (IPL) file.
‐ Configuration data elements, such as Prohibit Dynamic Connectivty Mask (PDCM),
Product Information Broker (PIB), port address names, mode register, and key counter.
‐ CUP configuration files.
• Ports moved into a logical switch will have persistent states based on the Active=Saved (ASM) bit
setting and other factors. Refer to Disabling and enabling ports - persistent states on page 48 for
more information.
Addressing modes
The following 256-area addressing modes provide extended functionality in dynamically created
partitions:
• Mode 0 - 10-bit addressing
• Mode 1 - zero-based addressing
• Mode 2 - port-based addressing
• The default switch is being used for FICON and no 48 or 64 port blades are installed in the chassis.
The default switch address mode 0 will be 8-bit area mode with zero based addressing (which
supports FICON) instead of 10-bit addressing.
• The chassis is not enabled for virtual fabrics and 48 or 64 port blades are not installed in the chassis.
NOTE
Although port addresses are assigned when moved to a logical switch, the address is not bound to the
port. To ensure that the link address for a control unit attached to a port does not change, the
portAddress command must be used to bind the address to the port even if you are not changing the
port address assignments. Refer to Sequential address binding on page 109 for a list of commands
that will bind ports on DCX or DCX 8510-8 Backbones fully populated with 32-port blades.
NOTE
To access all parameters controlled by the configure command, you must disable the switch. Any
configuration change to a logical switch that requires disabling the switch also requires disabling FMS
mode.
1. Disable FMS mode using the ficonCupSet fmsmode disable command.
2. Disable the switch using the switchDisable command.
3. Change the address mode using the configure command.
The configure command runs in interactive mode and presents you with a series of hierarchical
menus. To change the address mode, select the Fabric parameters menu, then select the Enable
a 256 Area Limit field and enter appropriate option values. For more information on this command
and an example of command output, refer to the Fabric OS Command Reference.
4. Enable the switch using the switchEnable command.
5. Enable FMS mode using the ficonCupSet fmsmode enable command.
The following tables compare the achievement of port persistent enable and disable states with
interactions between the ASM Mode setting, IPL state. and use of the portdisable or portenable
command.
TABLE 4 ASM 1 ON
Initial IPL state Blocked Blocked Unblocked Unblocked
frame header. High, medium, and low priority frames are allocated to different virtual channels (VCs).
High priority frames receive more VCs than medium priority frames, which receive more VCs than low
priority frames. The virtual channels are allocated according to the CS_CTL value.
NOTE
Check with your host and storage manufacturer to determine if they support Fibre Channel CS_CTL
prioritization on their devices.
The Fabric OS Administrator's Guide provides additional information on this feature and steps to
enable CS_CTL-based frame prioritization on your switch using the portCfgQos command.
The configureChassis command allows you to set chassis-level configuration parameters, such as
cfgload, custom, system, and Fabric OS attributes, to be used by ASIC drivers. The following
procedure uses the Fabric OS attribute, CS_CTL QoS Mode, to change CS_CTL values mapped to
VCs for priority levels from default to auto mode. While default mode maps CS_CTL values 1-8 (low),
9-16 (medium), and 17-24 (high) to specific VCs, auto mode maps CS_CTL values 1 (low), 2
(medium), and 3 (high) to VCs based on the ASIC supported by your platform. For DCX 8510-4, DCX
8510-8, and 6510 models, a single CS_CTL value is mapped to multiple VCs.
To enable auto mode for CS_CTL to VC mapping on a FICON switch, use the following steps.
NOTE
This process is disruptive as you must reboot the chassis to reset the configuration in the ASIC. You
cannot enable CS_CTL mapping on individual ports until this is done. If auto mode is selected, you
must select auto mode for each chassis in the fabric. Not doing so will mean different virtual channels
(VCs) will be used across the fabric with undesirable results.
1. Connect to the switch and log in to an account that has admin permissions.
2. Launch the configureChassis command.
3. Select no when prompted to modify cfgload attributes, custom attributes, and system attributes,
then select yes to modify Fabric OS attributes.
root> configurechassis
Configure...
cfgload attributes (yes, y, no, n): [no]
Custom attributes (yes, y, no, n): [no]
system attributes (yes, y, no, n): [no]
fos attributes (yes, y, no, n): [no] y
4. Select 1 (auto mode) when prompted to configure CS_CTL QoS mode.
CSCTL QoS Mode (0 = default; 1 = auto mode): (0..1) [0] 1
5. Reboot the switch or chassis.
6. To verify if auto mode or default mode is set for the chassis, enter the configShow --all command.
root> configshow -all | grep fos.csctlMode
fos.csctlMode:1
• If auto mode is set, the following should display in the command output:
fos.csctlMode:1
• If default mode is set, the following should display in the command output:
fos.csctlMode:0
To return to default mode, follow the preceding steps, but select 0 (default mode) for CS_CTL QoS
mode when modifying fos attributes through the configureChassis command.
NOTE
The configureChassis command does not provide options to enable the CS_CTL QoS mode on
individual ports in the chassis. Once you set the mode on the chassis, you must enable individual
ports to use this mode with the portCfgQos command. For details on the configureChassis and
portCfgQos commands, refer to the Fabric OS Command Reference for details.
Latency guideline
The maximum supported distance for a FICON channel is 300 Km (1.5 msec of delay). Synchronous
mirroring applications are generally limited to 100 Km (0.5 msec of delay). Greater distances require
that the FICON Acceleration feature be used with FCIP. The FICON Acceleration feature emulates
control unit response to the channel to make the devices appear closer to the channel than they
actually are.
NOTE
The CUP port address will always be xxFExx. Port addresses xxFE and xxFF are unavailable to assign
to physical ports when FMS is enabled. For more information, refer to Disabling ports 0xFE and 0xFF
on page 57.
FICON Management Server mode (FMS mode) must be enabled on the switch to enable CUP
management features. For more information, refer to FMS mode and FICON CUP on page 58.
The figure above is a simplified representation of a FICON environment and how CUP fits into that
environment. FICON switches with FMS enabled use port address "FE" as the embedded internal port
for host connectivity to the CUP, and port address "FF" is reserved. Therefore, addresses "FE" and
"FF" cannot be used for assignment to physical ports when FMS is enabled. For more information,
refer to Disabling ports 0xFE and 0xFF on page 57.
In a FICON environment, only one RMF logical partition (LPAR) should be configured to poll the CUP
for port statistical counter data to minimize contention for the CUP device. Host contention for access
to the CUP can lead to missing interrupts, potentially causing the host to box the device. It is a best
practice to have more than one CHPID defined as a path to the CUP to ensure redundancy for host-to-
CUP connectivity.
ATTENTION
Once FMS mode is enabled, you cannot bind addresses 0xFE or 0xFF to any ports using either the
portAddress or wwnAddress commands.
• The switch is advertised to the mainframe by the CUP as having a specific number of ports, which is
based on the hardware platform. The maximum supported switch size is 256 ports.
• The FE and FF addresses are not components of the Allow/Prohibit Matrixes (as they are called in
Brocade Network Advisor and Web Tools), because the Prohibit Dynamic Connectivity Masks
(PDCMs) for the ports associated with FE and FF are defined architecturally and cannot be modified
by the user.
• The port address name for the internal FE port can be read by the host, and corresponds to the
switch name that is set for the switch. The switch name can be modified by the host, with the
limitation that the host can write Extended Binary Coded Decimal Interchange Code (EBCDIC)
characters that cannot be converted directly to ASCII for display in the GUI or CLI management
interfaces. Such characters will be converted to dots (...) for these displays.
• Addresses FE and FF are not available for external connections. Before enabling FMS mode, you
must manually disable physical ports associated with FE and FF if there is an active device
connected to these ports. If there is an active device connected to either of these ports, Fabric OS
will not enable FMS mode until either the port is disabled or the device is taken offline (logs out of the
switch). If there is no active device connected, these ports will automatically disable when FMS
enables. Because the ports associated with FE and FF are not available after enabling FMS mode,
you should first move any fiber connected to either of them to another free port. For more
information, refer to Disabling ports 0xFE and 0xFF on page 57.
NOTE
Although 180 seconds should be the default value set on the switch when FMS mode is enabled, this
value may not be set when older versions of Fabric OS are upgraded. This value must be set on the
host system as well.
Command Description
ficonCupSet fmsmode Sets FICON Management Server mode on or off for the switch.
ficonCupSet MIHPTO Sets the Missing Interrupt Handler Primary Timeout value.
ficonCupSet modereg Sets the mode register bits for the switch.
ficonCupShow lp Displays the host logical path information for the CUP.
ficonCupShow fmsmode Displays the FICON Management Server enabled or disabled state for the switch.
ficonCupShow modereg Displays the mode register bit settings for the switch.
ATTENTION
If FMS mode is already enabled, disabling it may be disruptive to operations because ports that were
previously prohibited from communicating will be allowed to do so because prohibits are no longer
enforced.You must disable FMS mode to continue setting up CUP.
6. Install a CUP license on the switch. For more information on installing a license, refer to the Fabric
OS Administrator's Guide.
7. Be sure that the switch meets requirements for a high-integrity fabric, and then enable High-Integrity
Fabric (HIF) mode using the configure command. For details, refer to Meeting high-integrity fabric
requirements on page 35.
NOTE
You must enable HIF mode to enable FMS mode.
8. Enter the ficonCupSet modereg command to set the mode register bits. Refer to Mode register bit
settings on page 62 for more information.
9. Verify with the systems programmer that the CUP ports are defined properly.
FICON ports are defined as part of the sysgen process. The z/OS systems programmer can verify if
the CUP ports are defined properly.
On a Brocade DCX and DCX-4S Backbones, when Virtual Fabrics is not enabled or addressing Mode
1 is not used, you must disable physical ports associated with addresses 0xFE and 0xFF (if ports were
assigned by default) before enabling FMS mode. Note that physical ports associated with 0xFE and
0xFF may be different in logical switches using Mode 1 addressing. Disable these addresses manually
if there is an active device attached. If an active device is not attached to the ports, they will be
disabled automatically when you enable FMS mode.
You can only disable ports if there is a card present in the slot.
If FMS mode is enabled in a logical switch with Mode 2 addressing, the physical ports with addresses
0xFE and 0xFF are essentially wasted and may be used better elsewhere. This is the same for a
switch with Mode 1 addressing where all available addresses are used, and thus no addresses are
available for binding to these ports. On logical switches with Mode 1 addressing, such ports may be
reclaimed for external connections by binding a new address to the port, if unused addresses are
available.
Ports that are disabled for having addresses "FE or "FF" when FMS mode is enabled are not eligible
for port swapping.
NOTE
FMS mode cannot be enabled in a Virtual Fabrics environment when a logical switch is using Mode 0
addressing.
CAUTION
Use this procedure only under direction from technical support personnel. Indiscriminate use
can disrupt CCW processing and create errors at the host.
Entering the ficonCupSet fmsmode reset command does not change any configuration states, but
accomplishes the following:
• Resets the Linux mutexes (SMCB, CFG, Asy CCW)
• Resets the CUP logical path allegiance to neutral
• Clears pending sense data
• Clears pending CUE accept-status state
• Clears pending CUE-Owed
• You must manually disable ports associated with 0xFE and 0xFF using the portDisable command
before you can enable FMS mode, but only if there is an active device connected to these ports.
Otherwise, these ports will disable automatically when FMS is enabled. Refer to Disabling ports 0xFE
and 0xFF on page 57 for more information.
• Enabling FMS mode disables all ports with OxFE and OxFF addresses.
• If you enable FMS mode, you will get port-disable messages for physical ports with OxFE or OxFF
FICON addresses.
• You cannot enable FMS mode when physical ports have OxFE or OxFF FICON addresses with live
devices connected.
• Port swapping and blade swapping cannot be performed on ports where FMS mode is enabled and
ports have either OxFF or OxFE FICON addresses.
Upgrade considerations
Take the CUP offline before performing an upgrade or downgrade operation. The upgrade can disrupt
CUP exchanges, especially for active RMF sessions.
For Fabric OS 7.0 and later, a FICON CUP license is required to enable FMS mode. Consider the
following when using FMS mode and upgrading from Fabric OS v6.4.0 to v7.0 or later:
• If using Fabric OS v6.4.0 and FMS mode is enabled, but a FICON CUP license was never installed:
‐ You can manage the port connectivity configuration through management applications such
as Brocade Network Advisor and Web Tools.
‐ The In-band Management feature is not enabled.
‐ After a nondisruptive upgrade from Fabric OS v6.4.0 to v7.0.0 or later:
‐ You can manage the port connectivity configuration using management
applications such as Brocade Network Advisor and Web Tools. The Web Tools
mode register is accessible.
‐ To enable In-band Management, you must install a FICON CUP license and
disable and then re-enable FMS mode.
NOTE
A nondisruptive upgrade is not possible for the Brocade DCX 8510 Backbones, so the
nondisruptive upgrade information does not apply.
• If using Fabric OS v6.4 and FMS mode is enabled, but a FICON CUP license was installed and then
removed:
‐ You can manage the port connectivity configuration through management applications such
as Brocade Network Advisor and Web Tools. The FICON CUP Mode Register is accessible
through Web Tools.
‐ Inband Management is enabled, providing you do not disable FMS mode, disable and then
re-enable the switch, or power-cycle the switch.
• After a nondisruptive upgrade from Fabric OS v6.4.0 to v7.0.0 or later, the following events occur:
NOTE
The following only occur on platforms released prior to Fabric OS v7.0.0.
‐ You can manage port connectivity using the Allow/Prohibit Matrix in Brocade Network
Advisor and Web Tools. Note that these applications are not available unless you install a
FICON CUP license. The Mode Register is accessible via Web Tools and the Fabric OS
commands.
‐ In-band Management is enabled, provided you do not disable FMS mode, disable and then
enable the switch, or power-cycle the switch.
‐ If you disable FMS, a FICON CUP license is required to re-enable the feature.
• Upgrading with FMS mode enabled from Fabric OS v7.2.1 or prior to v7.3.0 or later requires that the
following attributes be configured. If FMS mode is enabled on the switch and these are not
configured, upgrade will fail.
‐ An insistent domain ID (IDID)
‐ A valid SCC_Policy (configured and activated)
‐ A Fabric Wide Consistency Policy => SCC:S (Strict mode)
NOTE
You enable the Fabric Wide Consistency Policy on the fabric once the switch joins the
fabric.
‐ HIF mode enabled.
ACP Alternate control prohibited. Because the Fabric OS CLI, Web Tools, and Brocade Network Advisor are
considered to be switch consoles, this bit has no effect on their operations. Attempts to set CUP
parameters through SNMP are denied when this bit is set on. The default setting is 1 (on).
ASM Active=Saved mode. When this bit is set on, all CUP configuration parameters are persistent, meaning
that they will be saved in nonvolatile storage in the IPL file that is applied upon a cold reboot or a power
cycle. The default setting is 1 (on).
DCAM Switch clock alert mode. When this bit is set on, a warning is issued when the date, tsClockServer, or
tsTimeZone commands are entered to set the time and date on the switch. The default setting is 0 (off).
HCP Host control prohibited. When this bit is set on, the host is not allowed to set CUP parameters. The
default setting is 0 (off).
POSC Programmed offline state control. If the bit is set (=1), the command from the host sets the switch offline.
If the bit is off (=0), the command from the host is rejected and the switch does not go offline. The default
setting is 1 (on).
UAM User alert mode. When this bit is set on, a warning is issued when an action is attempted that will write
CUP parameters on the switch. The default setting is 0 (off).
When the POSC bit=1, the FICON host is allowed to set the switch offline. The bit is set to 1 by default.
If you do not want to allow the host to have permission to set the switch offline, the bit should be turned
off. The state of this bit cannot be changed by the FICON host; it can only be changed through the
command line.
The ficonCupShow modereg command displays the mode register bit settings for the switch. A display
of 0 indicates that the mode register bit is set to off; 1 indicates that the bit is set to on. To display the
mode register bit HCP for the switch, enter the command as shown in the following example:
For more information on this command and an example of command output, refer to the Fabric OS
Command Reference.
The IPL is not replaced because Active=Saved mode is enabled. A warning message is displayed in
the event log to warn users that the IPL will not be overwritten.
*------------------------------------------------------------------
* Brocade Domain_ID=0x3c0000
*------------------------------------------------------------------
CNTLUNIT CUNUMBR=0D8,UNITADD=00,UNIT=2032,
PATH=(50,51),
LINK=(60FE,60FE)
IODEVICE ADDRESS=(0D8,1),CUNUMBR=0D8,UNIT=2032,STADET=Y,UNITADD=00
*---------------------------------------------------------------
CNTLUNIT CUNUMBR=0D9,UNITADD=00,UNIT=2032,
PATH=(8A,8B),
LINK=(22FE,22FE)
IODEVICE ADDRESS=(0D9,1),CUNUMBR=0D9,UNIT=2032,STADET=Y,UNITADD=00
*------------------------------------------------------------------
* Brocade Domain_ID=0x190000
*------------------------------------------------------------------
*
CNTLUNIT CUNUMBR=0DB,UNITADD=00,UNIT=2032,
PATH=(5A,5B),
LINK=(25FE,25FE)
IODEVICE ADDRESS=(0DB,1),CUNUMBR=0DB,UNIT=2032,STADET=Y,UNITADD=00
*
*------------------------------------------------------------------
Refer to Disabling and enabling ports - persistent states on page 48 for more information on enabling
and disabling ports and persistent states.
‐ Capacity 1
‐ Consumed 1
‐ Configured Blade Slots 1
This is a slot-based license for extension switches and blades.
• Advanced Extension (FTR_AE)
Required for multiple-circuit tunnels, FCIP Trunking, Adaptive Rate Limiting, and other FCIP
features. Advanced Extension is not required for FICON Acceleration features, but multiple circuits
could be used in a FICON emulation environment. This is a slot-based license for extension
switches and blades.
This license contains the following options:
‐ Capacity 1
‐ Consumed 1
‐ Configured Blade Slots 1
ATTENTION
Use the licenseShow command to verify the licenses are installed on the hardware at both ends of
the FCIP tunnel. For slot-based licenses, use the licenseSlotCfg --show command to display all slot-
based licenses in a chassis.
For more information on licensing, such as licensing required per product, refer to the Fabric OS
Software Licensing Guide.
The figure above shows how the primary volume and the secondary mirrored volume may be
geographically distant across an IP WAN. Updates to the primary disk volumes are completed by the
production applications and then staged in a control unit cache to be read by SDM. SDM then writes the
updates to the secondary volumes. The latency introduced by greater distance creates delays in
anticipated responses to certain commands. The FICON IU pacing mechanism may interpret delays as
an indication of a large data transfer that could monopolize a shared resource and react by throttling the
I/O. IBM z/OS Global Mirror emulation provides local responses to remote hosts, eliminating distance-
related delays. You can use the Brocade 7800 switch, Brocade 7840 switch, or an FX8-24 blade with
FICON emulation.
For information on configuring IBM z/OS Global Mirror emulation, refer to Configuring FICON emulation
on page 74. For information on displaying IBM z/OS Global Mirror emulation status and statistics,
refer to Displaying FICON emulation performance statistics on page 75.
Tape emulation
Tape emulation (also called tape pipelining) refers to the concept of maintaining a series of I/O
operations across a host-WAN-device environment and should not be confused with the normal FICON
streaming of CCWs and data in a single command chain. Normally, tape access methods can be
expected to read data sequentially until they reach the end-of-file delimiters (tape marks) or to write
data sequentially until either the data set is closed or an end-of-tape condition occurs (multi-volume file).
The emulation design strategy attempts to optimize performance for sequential reads and writes, while
accommodating any other nonconforming conditions in a lower performance non-emulating frame
shuttle. Because write operations can be expected to comprise the larger percentage of I/O operations
for tape devices (for archival purposes), they are addressed first.
For information on configuring tape emulation, refer to Configuring FICON emulation on page 74. For
information on displaying tape emulation status and statistics, refer to Displaying FICON emulation
performance statistics on page 75.
Printer emulation
Brocade extension switches and blades can operate with a Prizm FICON to ESCON converter and an
ESBT Bus/Tag Interface Module for Prizm from Optica Technologies, Inc. This allows FICON printer
emulation when connecting to printers such as the 1043, 3203, 3211, 3800, 4245, 4248, 6262, Xeros
Printer Access Facility (XPAF), and Intelligent Printer Data Stream (IPDS) models. Discovery of these
devices is automatic when FICON emulation is enabled on an FCIP tunnel.
For printer emulation, basic write streams are emulated to help resolve link latency resulting from
channel programming and associated FICON protocols. Because most supported printers will be
connected through the Prizm FICON to ESCON converter or ESBT Bus/Tag Interface Module, resolving
latency on the WAN is not required. The bottleneck will be the remote channel connection where it is
more important to have outstanding commands queued. Basic read commands known to be part of the
write command chain, typically those that read printer page counters, will be handled locally without
waiting for the actual read data from the device.
This emulation provides the following benefits:
• Optimized performance for FICON printers over distance when printers are extended through an
FCIP tunnel with FICON emulation.
• Intermediate ending status as appropriate to prevent FlCON channel protocol timeout (PTOV)
expirations.
• Acceptance of device status as required to prevent FICON control unit PTOV expirations.
• Prevention of "clutching" during print jobs.
For information on configuring printer emulation, refer to Configuring FICON emulation on page 74.
For information on displaying printer emulation status and statistics, refer to Displaying FICON
emulation performance statistics on page 75.
Teradata emulation
Teradata emulation reduces latency on links to Teradata Warehouse systems caused by WAN
propagation delays and bandwidth restrictions. It accomplishes this by processing selected FICON
commands for associated control, data, and status responses. FICON Teradata emulation is only
supported between FICON channels and FICON Teradata controllers.
For write commands, control and status frames are generated for the host side of the WAN in order to
pipeline write commands over the same or multiple exchanges. Such pipelined write commands and
their data are queued at the device side of the WAN for asynchronous transfer to the device.
For read operations received by the device side of the WAN, a number of anticipatory read commands
are autonomously generated and transferred to the device. The data and status associated with such
commands are sent to the host side of the WAN and queued in anticipation of host-generated read
commands.
FICON emulation for Teradata sequences over an FCIP tunnel is controlled by using FICON emulation
feature flags in the PortCfg fciptunnel command associated with the FCIP tunnel.
For information on configuring Teradata emulation or tape emulation, refer to Configuring FICON
emulation on page 74. For information on displaying Teradata emulation status and statistics, refer
to Displaying FICON emulation performance statistics on page 75.
• FICON networks with FCIP emulating and non-emulating tunnels do not support Dynamic Path
Selection (aptpolicy 3) configurations.
• When running FICON emulation features over an FCIP tunnel, it is highly recommended that both
switches providing the FCIP tunnel use the same Fabric OS release level.
ATTENTION
Make sure both ends of the tunnel match before bringing it up or the tunnel will not form.
1. Verify that the following ports are opened in the IP network for FCIP:
• TCP 3225--FC Class F frames
• TCP 3226--FC Class 2 and 3 frames
• TCP 3227--Brocade proprietary IP path performance utility (ipPerf)
• IP protocol 6--Transmission Control Protocol
• IP protocol 108--IP Payload Compression Protocol (IPComp)
2. Verify that the firmware is Fabric OS v6.1.0e or later using the firmwareShow command.
3. Verify the correct licenses are installed using the licenseShow command. You can verify all slot-
based licenses in a chassis using the licenseSlotCfg --show command.
4. Verify the switches have been set up according to the instructions in Configuring switched point-to-
point FICON on page 39 and Configuring cascaded FICON on page 43.
5. Configure the FCIP tunnel according to the Fabric OS FCIP Administrator's Guide.
6. Enable compression on the FCIP tunnel.
7. Configure FICON emulation on the tunnel that is solution-specific. Refer to Configuration examples
on page 74.
8. If you have more than one tunnel being configured, take the following actions:
• If there is more then one tunnel to the same destination and their tunnel commit rates are
different, configure the link cost values the same.
• Configure TI Zones ensuring that if there are multiple E_Ports of different link costs, from one
switch to another switch within TI Zones, configure the link cost values to be the same.
Configuration examples
You can configure FICON emulation with such features as Teradata read, Teradata write, printer
emulation, tape write pipelining, tape read pipelining, TIN/TIR, device level ACK, and IBM z/OS Global
Mirror (formerly XRC) emulation enabled.
Following are examples for enabling IBM z/OS Global Mirror, Teradata emulation, tape read and write
pipelining, and printer emulation.
Note that the preceding commands assume that an FCIP tunnel was a enabled for FICON operation
by using the -F or --ficon option. Following are two examples:
• To create a empty tunnel as a FICON FCIP tunnel, use the following command:
portcfg fciptunnel 1/17 create -F -c 1
• To modify a tunnel to be FICON enabled and with hardware compression enabled, use the following
command:
portcfg fciptunnel 1/17 modify -c 1 -F 1
ATTENTION
You must bring the FCIP tunnel down in order to make changes. The following process is disruptive to
all traffic on the tunnel.
1. Connect to the switches at both ends of the FCIP tunnel and log in using an account assigned to the
admin role.
2. Disable the FCIP tunnel at both ends where FICON emulation is enabled.
3. Use the portCfg fciptunnel modify command for the Brocade extension switches and blades.
4. Make matching FICON emulation changes to both ends of the tunnel.
5. Enable the FCIP tunnel.
VE HD HP DD DP LP CU DV
24 63 01 64 06 05 00 01
State A value of zero indicates emulation is idle. Any non-zero value indicates emulation
is active.
portshow xtun slot/ve_port ve_port -ficon -emul Statistics and status for printer, Teradata, tape,
and IBM z/OS Global Mirror (XRC) emulation.
portshow xtun slot/ve_port -ficon -emulxrc FICON IBM z/OS Global Mirror emulation
statistics and status.
portshow xtun slot/ve_port -ficon -emultape FICON tape emulation statistics and status.
portshow xtun slot/ve_port -ficon -emulprint FICON printer emulation statistics and status.
portshow xtun slot/ve_port -ficon -emultera FICON Teradata emulation statistics and status.
portshow xtun slot/ve_port -ficon -stats General statistics, including those for Teradata,
XRC, tape emulation, and FICON idle status
acceptance.
portshow xtun slot/ve_port -ficon -fdcb control block Teradata, tape, IBM z/OS Global Mirror, or
printer emulation statistics, including the device
extension of the FDCB for a specified Device
Control Block.
portshow xtun ve_port -ficon -tapeperf FICON tape pipelining performance statistics.
portshow xtun ve_port -ficon -xrcperf IBM z/OS Global Mirror performance statistics.
Refer to the portShow command section of the Fabric OS Command Reference for more information
on these commands.
• During the firmware download process, a failover to the backup processor occurs. Any outstanding
CUP commands are lost which results in a timeout. To avoid this, CUP should be varied offline
before starting a firmware download. The CUP may be varied back online after a successful
firmware download.
• All firmware downloads cause an I/O disruption when the new code is activated on the Brocade
7800 switch and FX8-24 blade. However, the Brocade 7840 switch supports the FCIP Hot Code
Load (HCL) feature which allows firmware downloads without I/O disruption.
NOTE
If upgrading multiple switches or blades, upgrade one at a time. This is not a requirement for the DCX
or DCX-4S platforms.
If FMS mode is enabled in a configuration file, but is disabled on the switch, the configDownload
command fails and displays an error message. This prevents undesirable conditions that could result
from enabling FMS mode on a switch that does not require it.
Fabric OS v6.2.0 or later is required for TI Zoning in a FICON fabric. For more information on TI
Zoning, refer to the Fabric OS Administrator's Guide.
NOTE
Session-based zoning enforcement is not recommended on a FICON switch. For more information on
session-based zoning enforcement, refer to the Fabric OS Administrator’s Guide.
D M=DEV(1002)
Response is:
Enhanced TI Zoning
Enhanced Traffic Isolation Zones allow you to specify a preferred and exclusive path through a
cascaded fabric for a particular flow by defining a set of devices or ports to appear in more than one
Enhanced TI Zone. A preferred path is one that has failover enabled and an exclusive path has
failover disabled . An exclusive path in the following topologies is beneficial as it limits the number of
hops between switches:
• Triangular
• System Data Mover
• Emulation
Enhanced TI Zoning for the three previously listed topologies is available only on the following
platforms running Fabric OS v6.4.0 or later:
• Brocade 5100
• Brocade 5300
• Brocade 7800
• Brocade DCX
• Brocade DCX-4S
Triangular topology
The triangular topology shown in the figure below is a common FICON deployment for managing data
center recovery functions. In this topology, a switch is located in each of three sites and connected to
the other two forming a triangle. In most environments, the purpose of this configuration is to give the
host application the ability to connect to storage media at the other two sites, which allows it to recover
from an outage at either site. Meanwhile, the storage systems constantly synchronize between the two
storage sites.
During normal operation, the routes from any device on one switch are a one-hop path to either of the
other two switches. However, if the path between any pair of switches fails, then a two-hop path is
available. Multi-hop paths are not certified for FICON configurations.
To get around the multiple-hop issue, use Enhanced TI Zoning to prohibit the connection between
E_Ports on the switches. For this setup you must define a device in two zones with failover disabled.
The figure below shows that the setup allows a source device to reach two separate destinations over
exclusive, one-hop paths, while preventing the source device from taking the two-hop path in the event
of a failure.
The preceding figure shows the following TI Zone definitions for the triangular topology:
• Blue Zone includes the channel, ISL 1, and CU A.
• Green Zone includes the channel, ISL 2, and CU B.
• Red Zone includes the CU A, ISL 3, and CU B.
Using these definitions with failover disabled, the topology achieves the desired behavior. The channel
can reach CU A only over ISL 1 and CU B only over ISL 2. If either ISL should fail, connectivity to the
destination with the failed ISL is prevented because the exclusive path has failed.
The host in the production site operates the SDM application that reads data from the DASD and writes
it to the remote tape. The path to the remote site is required to be isolated to support the bandwidth
requirements needed to write the data to the remote tape. The host in the backup site needs access to
the DASD in the production site to support a hot recovery scenario in the event the production host
goes down. Supporting these two functions requires the ability to define overlapping zones that isolate
the tape traffic from the DASD traffic and allow local access to the DASD ports.
The figure above shows the following Enhanced TI Zone definitions for the System Data Mover
topology:
• Green Zone includes Host A, ISL 1, and CU B.
• Orange Zone includes CU A, ISL 2, and Host B.
• Red Zone includes Host A and CU A.
• Blue Zone includes Host B and CU B.
Using these definitions with failover disabled achieves the desired behavior. The production channels
can access the local DASD and use an exclusive path to write data to the remote tape. The remote
channels can access the remote tape and use an exclusive path to write data to the production DASD.
In both cases, the production tape traffic is excluded from intruding on the backup DASD traffic and
vice versa.
Emulation topology
FICON device emulation topologies have unique requirements due to the characteristics of the
emulation devices. Each device must track the emulated sessions and maintain the state of the
emulated devices in order to satisfy both ends of the connection. Because this is done over very long
distances, care must be taken when configuring the ISL connections to insure that the sessions remain
along exclusive paths. By maintaining an exclusive path between the emulation units, the recovery
scenarios become manageable and allow system-level mechanisms to take precedence, such as
device level recovery at the channel or control unit.
Generally, channel-to-control unit access is deployed as a one-to-many topology where one channel is
defined to access many control units. FICON environments, however, are deployed in multiple
redundant configurations that also define access from one control unit to many channels. When
emulation is involved, these configurations can result in overlapping channel and control unit
connections that make emulation isolation difficult. By using Enhanced TI Zoning to support device
membership in multiple zones, these types of emulation configurations are addressed.
The preceding figure shows the following Enhanced TI Zoning definitions for the emulation topology:
• Green TI Zone includes channel 0, ISL 80, ISL 81, Grid 1, and Grid 4.
• Blue TI Zone includes channel 1, ISL 82, ISL 83, and Grid 4.
The unique characteristic of these zones is that each channel is defined with two control unit device
ports and one of the control unit device ports, Grid 4, is in both zones. In a traditional TI Zone
definition, these zones would have to be combined to provide the appropriate connectivity, but in the
event of an ISL failure, this configuration would lead to a multi-hop, multi-emulation path that would
severely affect performance. Multi-hop paths are not certified for FICON configurations.
Port fencing
Occasionally, bad optics and cables can cause errors to occur at such a rapid rate that error processing
and RSCN transmission and processing results in fabric performance problems. Port fencing allows you
to limit the number of errors a port can receive by forcing a port offline when certain error thresholds are
met.
The port fencing feature is configured through Fabric Watch. For more information on configuring port
fencing, refer to the Fabric Watch Administrator's Guide.
NOTE
Establish a Telnet session with a tool such as PuTTy that allows the columns to be increased. This is
because some of the displays use more than the standard 80 columns that programs such as
HyperTerminal support. The recommended number of columns is 120.
FICON information
You can display link incidents, registered listeners, node identification data, and FRU failures, as
described in the following sections.
Link incidents
The Registered Link Incident Report (RLIR) ELS contains the link incident information sent to a
listener N_Port.
To display link incidents, connect to the switch, log in as user, and enter one of the following
commands:
• For the local switch: ficonShow rlir
• For all switches defined in the fabric: ficonShow rlirrlir fabric
Identifying ports
The ficonShow rlir command displays, among other information, a tag field for the switch port. You
can use this tag to identify the port on which a FICON link incident occurred. The tag field is a
concatenation of the switch domain ID and port number, in hexadecimal format. The following example
shows the top portion of command output which shows a link incident for the switch port at domain ID
120, port 93 (785d in hex):
{
{Fmt Type PID Port Incident Count TS Format Time Stamp
0x18 F 785d00 93 1 Time server Thu Apr 22 09:13:32 2004
Port Status: Link not operational
Link Failure Type: Loss of signal or synchronization
Registered Port WWN Registered Node WWN Flag Node Parameters
50:05:07:64:01:40:16:03 50:05:07:64:00:c1:69:ca 0x10 0x200115
(output truncated)
For complete examples of output from this command, refer to the ficonShow command section of the
Fabric OS Command Reference.
Registered listeners
To display registered listeners for link incidents, connect to the switch, log in as user, and enter one of
the following commands:
• For the local switch: ficonShow lirr
• For all switches defined in the fabric: ficonShow lirr fabric
For an example of local Link Incident Record Registration (LIRR) database output, refer to the
ficonShow command section of the Fabric OS Command Reference.
When you configure an LIRR entry and it is current, it is marked with -C. If the Management Server
marks an entry as current based on order, it is marked with -C. If the Management Server marks an
entry as secondary, it is marked with -S. Following is an example of output with the -C and -S flags.
Switch Port WWN Listener Type
20:2d:00:60:69:80:1e:4e Conditional-C
20:0d:00:60:69:80:1e:4f Conditional-S
If the LIRR is configured, then the a valid port number displays as shown in the following example.
Current LIRR device port number: 27 (0x1b)
Error data
The following examples show error data available from the host. The number identifying the failed
Brocade FRU (refer to FRU numbers on page 92) is derived from a single bit from byte 29 as the high
order bit and four bits from byte 30 or 31 as the low order bits (refer to the data in bold in the second
lines of the following examples). The FRU numbers are decoded for display as the FRU names, which
display in the third lines of the following examples.
FRU numbers
The following Brocade FRU numbers are provided for error reports and associated FRU names that
display in error reports. Note that the FRU number does not display in the error reports, but its
associated FRU name does display.
• 1 = SLOT1
• 2 = SLOT2
• 3 = SLOT3
• 4 = SLOT4
• 5 = SLOT5
• 6 = SLOT6
• 7 = SLOT7
• 8 = SLOT8
• 9 = SLOT9
• 10 = SLOT10
• 11 = SLOT11
• 12 = SLOT12
• 17 = PWR1
• 18 = PWR2
• 19 = PWR3
• 20 = PWR4
• 24 = FAN1
• 25 = FAN2
• 26 = FAN3
• 30 = CHASSIS
• 31 = WWN-CRD
Important notes
Consider the following when swapping port area IDs:
• Ports that have been swapped cannot be moved to another logical switch.
• Ports with area OxFE or OxFF addresses cannot be swapped when FMS mode is enabled.
• You can use the portSwapShow command to display information about swapped ports in a switch.
• You can use the portSwap command to disable the port swapping feature. You cannot use the
portSwap command after this feature is disabled. The enabled state of the port swapping feature is
persistent across reboots and power cycles. Enabling and disabling the port swapping feature does
not affect previously executed port swapping operations.
Blade swapping
Blade swapping allows you to swap one blade with another of the same type; in this way, you can
perform a FRU replacement with minimal traffic disruption. The entire operation is accomplished when
the bladeSwap command runs on Fabric OS. Fabric OS then validates each command before actually
implementing the command on the platform. If an error is encountered, the blade swap quits without
disrupting traffic flowing through the blades. If an unforeseen error does occur during the running of the
bladeSwap command, an entry will be made into the RASlog and all ports that have been swapped as
part of the blade swap operation will be swapped back. On successful completion of the command, the
source and destination blades are left in a disabled state, allowing you to complete the cable move.
Blade swapping is based on port swapping and has the same restrictions:
• Shared area ports cannot be swapped.
• Faulty blades cannot be swapped.
• Ports or blades with area OxFE or OxFF addresses cannot be swapped when FMS mode is enabled.
• Ports that are part of a trunk group cannot be swapped.
• GbE ports cannot be swapped.
• Swapping ports between different logical switches is not supported. The ports on the source and
destination blades must be in the same logical switch.
• Undetermined board types cannot be swapped. For example, a blade swap will fail if the blade type
cannot be identified.
• Blade swapping is not supported when swapping to a different model of blade or a different port
count. For example, you cannot swap an FC8-32 blade with an FC8-48 port blade.
• Blade swapping is not supported on application blades.
• Blade swapping is supported on the FX8-24 blade, but only the 12 Fibre Channel ports will swap.
The GbE and VE_ Ports will not be swapped.
For procedures and more information on blade swapping, refer to the Fabric OS Administrator's Guide.
Cascading: No path - Entry Link is ".." message This typically occurs when one or more of the required
high integrity features for cascading have not been met.
Check the following:
• Review each step in Configuring switched point-to-
point FICON on page 39 and Configuring cascaded
FICON on page 43 and validate that all features are
configured properly.
• Validate that CUP is online for each backbone or
switch in the fabric with the data center’s system
programmer.
Packets are being dropped between two FICON units. Distance between devices is impacting data.
When planning cable needs, the following criteria must
be considered:
• Distance considerations
• Fiber Optic Sub Assembly (FOSA) type (SW or LW)
• Cable specifications (SM or MM)
• Patch Panel Connections between FOSA ports (link
loss .3-5 dB per)
• Maximum allowable link budget (dB) loss
For cabling, the most important factor of a Fibre
Channel link is the selection of the Fibre Optic Sub
Assembly (FOSA) and matching cable type, to support
the required distance. Both ends of the optical link must
have the matching FOSA (SFP) types.
Channel path with two-byte addressing does not come The fabric may not be configured properly, cable
online. connections are not correct, or cable is wrong type.
Find out what CHPID the operator is attempting to bring
online. In FICON environments, only the egress ports
are defined in the IOCDS. The CHPID number of the
channel appears in the node list, but only if the channel
path has logged in. If the channel path will not come
online, it may not have completed a Fibre Channel
fabric login. If this is the case, you will have to rely on
customer documentation to determine the switch port to
which the channel is connected.
Next, find out what type of error message the operator
is getting. An IOS error message is reported when there
is an error external to the host; an IEE error message
indicates something is wrong internal to the host.
Fabric Not Configured Properly
• Make sure that the insistent domain ID is set and the
domain IDs are unique for all switches in the fabric.
• Make sure that the same binding method is used on
all switches in the fabric.
• Scroll through the node list and make sure there are
no invalid attachments.
• Make sure that the re-routing delay is turned off.
Cabling
Validate that cables are connected where intended.
When using multimode fiber, make sure that all fiber is
either all 50u or all 62.5u. A common mistake is to mix
50u and 62.5u fiber cables, which causes errors at
patch panels.
Example of a Boxed Channel
In the following example, the CHPIDs with two-byte
addressing did not come online. In response, the
CHPID was configured offline and then reconfigured
back online.
MVS console commands and responses (lines
beginning with "IEE" are responses and all other lines
are entries).
CF CHP(60),OFFLINE
IEE503I CHP(60),OFFLINE
IEE712I CONFIG PROCESSING COMPLETE
CF CHP(60),ONLINE
IEE754I NOT ALL PATHS BROUGHT ONLINE
WITH CHP(60)
IEE502I CHP(60),ONLINE
IEE712I CONFIG PROCESSING COMPLETE
The following line displays the status of the CHPID with
the "D M =" command. All other lines are responses.
D M=CHP(60)
IEE174I 03.29.45 DISPLAY M 021
CHPID 60: TYPE=1D, DESC=FICON
INCOMPLETE, ONLINE
DEVICE STATUS FOR CHANNEL PATH 60
0 1 2 3 4 5 6 7 8 9 A B C D E F
006 . $@ . . . . . . . . . . . . . .
SWITCH DEVICE NUMBER = NONE
The Control Unit Port cannot access the switch. Domain or port ID has changed on CHPID link.
A two-byte CHPID link is defined using a domain and
port ID that must remain consistent. Any change in the
physical link, such as a change in a domain or port ID,
will prevent storage control unit access.
Use the configure command to verify and set the
insistent domain ID (IDID) parameter.
FICON:admin> configure
For more information and examples of using this
command to set the IDID parameter, refer to the
Configure command in the Fabric OS Command
Reference.
Unable to "vary online" FICON CUP port on the switch . FICON CUP port cannot go online.
Use the haFailover command on the Backbone. This
the only known fix as there is no known firmware
solution.
Troubleshooting FICON
This section provides information-gathering and troubleshooting techniques necessary to fix your
problem.
‐ ficonCupShow fmsmode
‐ ficonCupShow modereg
‐ ficonDbg dump rnid
‐ ficonDbg log
‐ ficonShow lirr
‐ ficonShow rlir
‐ ficonShow rnid
‐ ficonShow switchrnid
‐ ficuCmd dump -A
• The type of mainframe involved including the make, model, and driver levels in use.
• The type of storage array installed. Many arrays emulate a certain type of IBM array and the make,
model, and firmware of the array in use must be provided.
• Other detailed information for protocol-specific problems:
‐ Port data structures, displayed using the ptDataShow command.
‐ Port registers, displayed using the ptRegShow command.
The following actions must be performed to troubleshoot all FICON setups:
• Use the configUpload command to gather configuration files. Provide the IOCDS mainframe file.
This will define how all mainframe ports are configured.
• Verify that Dynamic Load Sharing (DLS) has been disabled with the dlsReset command.
If DLS is enabled, traffic on existing ISL ports may be affected when one or more new ISLs is added
between the same two switches. Specifically, adding the new ISL may result in dropped frames as
routes are adjusted to take advantage of the bandwidth provided. By disabling DLS, you ensure that
there will be no dropped frames. In thesupportShow output, search for route.stickyRoutes and
check for a value of 1 or enter the dlsShow command.
• Verify that IOD is enabled using the iodSet command to ensure in-order delivery.
In the output from the supportShow or supportSave, search for route.delayReroute and check for
a value of 1 as this indicates that the feature is turned on. Alternately, you can enter the iodShow
command.
NOTE
There is no requirement to have a secure fabric in a switched point-to-point topology.
CUP diagnostics
FICON CUP supports the IBM Health Checker for zOS to diagnose problems. Refer to the IBM
Redbooks website at www.redbooks.ibm.com for details on using the IBM Health Checker for z/OS.
• Ensure no device is plugged into port 254 on the Brocade DCX Backbone.
• Verify with the systems programmer that the CUP ports are defined properly.
FICON ports are defined as part of the sysgen process. The z/OS systems programmer can verify if
the CUP ports are defined properly.
• Verify that ports 254 and 255 display Disabled (FMS Mode) by entering the switchShow command.
If not, enter the portDisable command on the appropriate models for the related ports:
‐ On a Brocade 48000, disable 10/30 and 10/31.
‐ On a Brocade DCX Backbone, disable 12/30 and 12/31.
Troubleshooting NPIV
Gather all information as listed in General information to gather for all cases on page 97 and Gathering
additional information on page 99.
Be sure that NPIV licenses are installed if using Fabric OS v5.0.x. There is no license requirement for
Fabric OS v5.1.0 and later.
● Introduction................................................................................................................... 101
● Supported platforms with end-of-support announcements........................................... 101
● Currently supported platforms.......................................................................................102
● Supported Brocade blades............................................................................................103
Introduction
This appendix provides the following support details:
• Brocade products with end-of-support announcements supported for FICON for different IBM
systems and drivers (Supported platforms with end-of-support announcements on page 101).
• Brocade products currently supported for FICON for different IBM systems and drivers (Currently
supported platforms on page 102).
• Brocade blades supported for FICON in various Brocade Switch and Data Center Backbone products
(Supported Brocade blades on page 103).
Due to the nature of environments in which the mainframes are used, product qualification for FICON is
more rigorous than with other products. As a result, not all product combinations and versions of Fabric
OS are qualified for FICON.
z900 & z800 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0
Driver 3GF
z990 & z890 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0 v7.3.0
Driver 55K
TABLE 12 Currently supported blades in Brocade Switch and Data Center Backbone products
1. The Ethernet ports of an FX8-24 extension blade can operate with another FX8-24 or Brocade 7800
switch. The maximum number of FX8-24 blades supported in a chassis is four. An FX8-24 can
operate with an FX8-24E; however, IPsec and compression is only supported on one 10GbE port on
the FX8-24.
2. The 64-port card is not supported for FICON but may be in the same chassis used for FICON. It must
be in a logical switch that does not have any FICON traffic.
3. For the 48-port card in an 8-slot chassis, FICON is supported in a logical switch configured for zero-
based addressing mode. For this card in a DCX-4S Backbone, FICON is supported only on a logical
switch defined for port-based addressing (mode 2).
4. FC10-6 10 Gbps port blade for ISL connections.
Unsupported blades
The following port blades are not supported in a FICON environment:
• FCOE10-24
• FS8-18
• FC4-48
• FC8-32E
• FC8-48E
This appendix provides basic steps and commands to quickly configure a switch for fabric and possible
FICON and cascaded FICON operation.
For detailed concepts, procedures, and additional Fabric OS command options to configure a switch for
point-to-point and cascaded FICON operation, refer to Administering FICON Fabrics.
The following tables are included in this appendix:
• The Switch configuration example and commands table provides the basic steps and commands in a
checklist format to quickly configure a switch for fabric and possible FICON operation.
• The Cascading (2-byte addressing) example and commands table provides additional steps and
commands for FICON cascading (two-byte addressing).
Create all logical switches and add all intended ports to each logical lscfg
switch.
Set in-order delivery with lossless DLS enabled. iodset --enable -losslessDLS
Set the principal switch. This can only be one switch in the fabric. In fabricprincipal 1
mixed M-Type and B-Type fabrics, the M-Type must be the principal
switch.
If you are enabling FMS (CUP), disable all ports assigned address portdisable
0xFE and 0xFF.
NOTE
High-integrity fabric requirements must be met and HIF mode
enabled to enable FMS mode.
Configure zones. This is much easier with Brocade Network Advisor cfgclear
or Web Tools. Typically, scripts are used when using Fabric OS
commands. cfgcreate
cfgdelete
cfgdisable
cfgenable
cfgremove
cfgsave
cfgshow
Set port fencing. Setting port fencing parameters is much easier with To clear alarms:
Brocade Network Advisor.
fwalarmsfilterset 0
To define alarms:
fwconfigure
Respond to menu driven prompts
accordingly.
To enable the defined alarms:
fwalarmsfilterset 1
Parameters in the table below are only required when two-byte link addressing is used in the IOCP.
Normally, these parameters are not set when single-byte addressing is used.
Feature Parameter
Before binding an address with the portAddress --bind command, disable the port with the
portDisable command. To change the bound address, you must first unbind the addresses. Refer to
Unbinding multiple ports on page 118 for an easy method to unbind multiple ports at once.
For more information on port addressing modes, refer to Addressing modes on page 46.
date
# Slot 1
portaddress --bind 1/0 0000
portaddress --bind 1/1 0100
portaddress --bind 1/2 0200
portaddress --bind 1/3 0300
portaddress --bind 1/4 0400
portaddress --bind 1/5 0500
portaddress --bind 1/6 0600
portaddress --bind 1/7 0700
portaddress --bind 1/8 0800
portaddress --bind 1/9 0900
portaddress --bind 1/10 0A00
portaddress --bind 1/11 0B00
portaddress --bind 1/12 0C00
portaddress --bind 1/13 0D00
portaddress --bind 1/14 0E00
portaddress --bind 1/15 0F00
portaddress --bind 1/16 1000
portaddress --bind 1/17 1100
portaddress --bind 1/18 1200
portaddress --bind 1/19 1300
portaddress --bind 1/20 1400
portaddress --bind 1/21 1500
portaddress --bind 1/22 1600
portaddress --bind 1/23 1700
portaddress --bind 1/24 1800
portaddress --bind 1/25 1900
portaddress --bind 1/26 1A00
portaddress --bind 1/27 1B00
portaddress --bind 1/28 1C00
portaddress --bind 1/29 1D00
portaddress --bind 1/30 1E00
portaddress --bind 1/31 1F00
# Slot 2
portaddress --bind 2/0 2000
portaddress --bind 2/1 2100
portaddress --bind 2/2 2200
portaddress --bind 2/3 2300
portaddress --bind 2/4 2400
portaddress --bind 2/5 2500
portaddress --bind 2/6 2600
portaddress --bind 2/7 2700
portaddress --bind 2/8 2800
portaddress --bind 2/9 2900
portaddress --bind 2/10 2A00
portaddress --bind 2/11 2B00
portaddress --bind 2/12 2C00
portaddress --bind 2/13 2D00
portaddress --bind 2/14 2E00
portaddress --bind 2/15 2F00
portaddress --bind 2/16 3000
portaddress --bind 2/17 3100
portaddress --bind 2/18 3200
portaddress --bind 2/19 3300
portaddress --bind 2/20 3400
portaddress --bind 2/21 3500
portaddress --bind 2/22 3600
portaddress --bind 2/23 3700
portaddress --bind 2/24 3800
portaddress --bind 2/25 3900
portaddress --bind 2/26 3A00
portaddress --bind 2/27 3B00
portaddress --bind 2/28 3C00
portaddress --bind 2/29 3D00
portaddress --bind 2/30 3E00
portaddress --bind 2/31 3F00
# Slot 3
portaddress --bind 3/0 4000
portaddress --bind 3/1 4100
portaddress --bind 3/2 4200
portaddress --bind 3/3 4300
portaddress --bind 3/4 4400
portaddress --bind 3/5 4500
portaddress --bind 3/6 4600
portaddress --bind 3/7 4700
portaddress --bind 3/8 4800
portaddress --bind 3/9 4900
portaddress --bind 3/10 4A00
portaddress --bind 3/11 4B00
portaddress --bind 3/12 4C00
portaddress --bind 3/13 4D00
portaddress --bind 3/14 4E00
portaddress --bind 3/15 4F00
portaddress --bind 3/16 5000
portaddress --bind 3/17 5100
portaddress --bind 3/18 5200
portaddress --bind 3/19 5300
portaddress --bind 3/20 5400
portaddress --bind 3/21 5500
portaddress --bind 3/22 5600
portaddress --bind 3/23 5700
portaddress --bind 3/24 5800
portaddress --bind 3/25 5900
portaddress --bind 3/26 5A00
portaddress --bind 3/27 5B00
portaddress --bind 3/28 5C00
portaddress --bind 3/29 5D00
portaddress --bind 3/30 5E00
portaddress --bind 3/31 5F00
# Slot 4
portaddress --bind 4/0 6000
portaddress --bind 4/1 6100
portaddress --bind 4/2 6200
portaddress --bind 4/3 6300
portaddress --bind 4/4 6400
portaddress --bind 4/5 6500
portaddress --bind 4/6 6600
portaddress --bind 4/7 6700
portaddress --bind 4/8 6800
portaddress --bind 4/9 6900
portaddress --bind 4/10 6A00
portaddress --bind 4/11 6B00
portaddress --bind 4/12 6C00
portaddress --bind 4/13 6D00
portaddress --bind 4/14 6E00
portaddress --bind 4/15 6F00
portaddress --bind 4/16 7000
portaddress --bind 4/17 7100
portaddress --bind 4/18 7200
portaddress --bind 4/19 7300
portaddress --bind 4/20 7400
portaddress --bind 4/21 7500
portaddress --bind 4/22 7600
portaddress --bind 4/23 7700
portaddress --bind 4/24 7800
portaddress --bind 4/25 7900
portaddress --bind 4/26 7A00
portaddress --bind 4/27 7B00
portaddress --bind 4/28 7C00
portaddress --bind 4/29 7D00
portaddress --bind 4/30 7E00
portaddress --bind 4/31 7F00
# Slot 9
portaddress --bind 9/0 8000
portaddress --bind 9/1 8100
portaddress --bind 9/2 8200
portaddress --bind 9/3 8300
portaddress --bind 9/4 8400
portaddress --bind 9/5 8500
portaddress --bind 9/6 8600
portaddress --bind 9/7 8700
portaddress --bind 9/8 8800
portaddress --bind 9/9 8900
portaddress --bind 9/10 8A00
Before binding an address with the portAddress --bind command, disable the port with the
portDisable command. To change the bound address, you must first unbind the addresses. Refer to
Unbinding multiple ports on page 118 for an easy method to unbind multiple ports at once.
For more information on port addressing modes, refer to Addressing modes on page 46.
date
# Slot 1
portaddress --bind 1/0 0000
portaddress --bind 1/1 0100
portaddress --bind 1/2 0200
portaddress --bind 1/3 0300
portaddress --bind 1/4 0400
portaddress --bind 1/5 0500
portaddress --bind 1/6 0600
portaddress --bind 1/7 0700
portaddress --bind 1/8 0800
portaddress --bind 1/9 0900
portaddress --bind 1/10 0A00
portaddress --bind 1/11 0B00
portaddress --bind 1/12 0C00
portaddress --bind 1/13 0D00
portaddress --bind 1/14 0E00
portaddress --bind 1/15 0F00
portaddress --bind 1/16 8000
portaddress --bind 1/17 8100
portaddress --bind 1/18 8200
portaddress --bind 1/19 8300
portaddress --bind 1/20 8400
portaddress --bind 1/21 8500
You can use the following worksheet for recording FICON configuration information.
NOTE
Attached E_Ports can be for an ISL, control unit (CU) interface, or central processor complex (CPC)
channel path identifier (CHPID).
HCD Defined Switch ID _________(Switch ID) Cascaded Directors No___ Yes ___
FICON® Switch Domain ID _________ (Switch @) Corresponding Cascaded Switch Domain ID ____
Fabric Name _____________________
Slot Port Port Laser Port Node Machine Model Serial ISL, CU
Number Number Address Type: Name Type Type Number I/F, CPC
LX / SX CU / CHPID
CHNL
Extended Binary Coded Decimal Interchange Code (EBCDIC) is an 8-bit character encoding (code
page) used on IBM mainframe operating systems such as z/OS and S/390. Code page 37 is an
EBCDIC code page with full Latin-1 character set.
H P
High-Integrity Fabric Mode 37 packets being dropped 94
High-Integrity Fabric requirements 35 persistently enabling/disabling ports 66
persistent port states 48
platforms supported for FICON 102
I port and switch naming standards 56
port disable and enable states 48
IBM z/OS Global Mirror emulation 68 port swap limitations
identifying ports from the tag fieldtag field, interpreting virtual fabrics 49
90 port swapping 93
IDID 13 printer emulation 71
insistent domain ID 13
intermix mode 13
invalid attachment/isolated state 35 Q
QSA requirements 35
L
latency guidelines 14, 52 R
link and FC addressing 32
link incidents 30 recording configuration information 64
link incidents, displaying 30, 90 registered listeners 30, 91
LIRR 30 RLIR
clearing the database 30
link incidents, displaying 30
RNID node identification 30
routing policy 45
T
tape emulation 69
Tape Pipelining 74
Teradata emulation 72
Teradata statistics 77
triangular topology 83
troubleshooting NPIV 100
U
unable to ’vary online’ 94
upgrade considerations 79
X
XISL 45
XRC 68
Z
zoning and PDCM considerations 30