MQ: First-failure support technology (FFST)

First-failure support technology (FFST)

Overview

For WebSphere MQ for UNIX systems, FFST information is recorded in a file in the /var/mqm/errors directory.
These errors are normally severe, unrecoverable errors, and indicate either a configuration problem with the system or a WebSphere MQ internal error.
The files are named AMQnnnnn.mm.FDC, where:

nnnnn Is the ID of the process reporting the error

mm Is a sequence number, normally 0

When a process creates an FFST record, it also sends a record to syslog. The record contains the name of the FFST file to assist in automatic problem tracking.
The syslog entry is made at the user.error level. See the operating-system documentation about syslog.conf for information about configuring this.
Some typical FFST data is shown in Figure 28.
Figure 28. FFST report for WebSphere MQ for UNIX systems
+-----------------------------------------------------------------------------+
|                                                                             |
| WebSphere MQ First Failure Symptom Report                                   |
| =========================================                                   |
|                                                                             |
| Date/Time         :- Friday March 15 17:56:51 SGT 2002                      |
| Host Name         :- sunrts3 (SunOS 5.7)                                    |
| PIDS              :- 5724B4102                                              |
| LVLS              :- 530                                                    |
| Product Long Name :- WebSphere MQ for Sun Solaris                           |
| Vendor            :- IBM                                                    |
| Probe Id          :- RM161000                                               |
| Application Name  :- MQM                                                    |
| Component         :- rrmChangeClq                                           |
| Build Date        :- Mar 13 2002                                            |
| CMVC level        :- p000-L020312                                           |
| Build Type        :- IKAP - (Production)                                    |
| UserID            :- 00001001 (mqm)                                         |
| Program Name      :- amqrrmfa                                               |
| Process           :- 00019454                                               |
| Thread            :- 00000001                                               |
| QueueManager      :- REGR                                                   |
| Major Errorcode   :- rrcE_CLUS_COMMAND_ERROR                                |
| Minor Errorcode   :- OK                                                     |
| Probe Type        :- MSGAMQ9413                                             |
| Probe Severity    :- 2                                                      |
| Probe Description :- AMQ9413: Repository command format error, command code |
|   0                                                                         |
| FDCSequenceNumber :- 0                                                      |
|                                                                             |
+-----------------------------------------------------------------------------+
 
MQM Function Stack
rrmProcessMsg
rrmChangeClq
xcsFFST
 
MQM Trace History
---{ zstVerifyPCD
---} zstVerifyPCD rc=OK
---{ ziiMQCMIT
----{ ziiCreateIPCCMessage
-----{ zcpCreateMessage
-----} zcpCreateMessage rc=OK
----} ziiCreateIPCCMessage rc=OK
----{ ziiSendReceiveAgent
-----{ zcpSendOnPipe
 
...
 
The Function Stack and Trace History are used by IBM to assist in problem determination. In most cases there is little that the system administrator can do when an FFST report is generated, apart from raising problems through the IBM Support Center.
However, there are some problems that the system administrator might be able to solve. If the FFST shows out of resource or out of space on device descriptions when calling one of the IPC functions (for example, semop or shmget), it is likely that the relevant kernel parameter limit has been exceeded.
If the FFST report shows a problem with setitimer, it is likely that a change to the kernel timer parameters is needed.
To resolve these problems, increase the IPC limits, rebuild the kernel, and restart the machine.

WebSphere is a trademark of the IBM Corporation in the United States, other countries, or both.

IBM is a trademark of the IBM Corporation in the United States, other countries, or both.

nnnnn	Is the ID of the process reporting the error
mm	Is a sequence number, normally 0