Troubleshooting
Problem
Your application that previously worked receives MQRC 2085 when it does an MQOPEN for an existing cluster queue.
Symptom
The application receives reason code 2085 (0x00000825 MQRC_UNKNOWN_OBJECT_NAME).
Other possible symptoms include:
- CSQX412E CSQXREPO Misdirected repository command,
target CSQ1.BE55F328BB1CE760
sender QM1_2005-02-16_10.56.51
where the target QMID does not match the current QMID of the queue manager - CSQX431I CSQXREPO Repository unavailable,
cluster CLUSTER1,
channel TO.CSQ1,
sender CSQ2.BDAA5ABACF0EB081 - 2082 (X'0822') MQRC_UNKNOWN_ALIAS_BASE_Q
- CSQX053E CSQXFFST Error information recorded in CSQSNAP data set
Eyecatchers in the FFST entry in CSQSNAP include
* X...XFFSrrmChangeClqMgr.*
* ....Duplicate CLQMGR....*
* ....Exis*
*ting UUID...<qmid1>.............*
* ....Dupl*
*icate UUID..<qmid2>.............*
The Existing UUID and Duplicate UUID data areas include eyecatchers with QMIDs. See Additional information below for a description of the format of the QMID.
Cause
DISPLAY CLUSQMGR(*) QMID and/or error messages show that there is more than one unique QMID for a given queue manager, for example:
- CSQ1.B6CA230882A37602
CSQ1.BE55F328BB1CE760
1. The queue manager might have been deleted and then re-created and redefined.
2. It might have been cold-started on z/OS, without first following the procedure to remove a queue manager from a cluster.
Reference the topics DISPLAY CLUSQMGR displays a queue manager twice and RESET CLUSTER in the WebSphere MQ for z/OS Queue Manager Clusters manual.
Resolving The Problem
- On the queue manager that is listed in the duplicate QMIDs, issue
display qmgr qmid
to find the "good" QMID.
- Remove reference to the "bad" QMID (the one with the older date/time) from the cluster by issuing
reset cluster(clustername) qmid(qmid) action(FORCEREMOVE) queues(yes)
from a full repository.
- For distributed platforms, issue these commands in runmqsc or using the MQ Explorer from Windows.
- For the z/OS platform, issue the command in the joblog preceded by a slash and the cpf (command prefix) for the queue manager subsystem, for example: /+csq1 display...
As stated in the topic A queue manager does not rejoin the cluster, the CLUSSDR on the queue manager that was RESET might be STOPPED. If so, manually START it.
If the reset command does not appear to be actioned, see if the SYSTEM.CLUSTER.COMMAND.QUEUE is disabled or if messages are built up on it. If so, the repository manager has had a problem--look for related error messages in the CHIN joblog.
Additional information
A QMID is an internally generated identifier that consists of the queue manger name and a date stamp whose format depends on the platform.
- For a z/OS queue manager, the QMID includes a 16-digit STCK timestamp, for example
- For a distributed queue manager an example of a QMID is
CSQ1.B6CA230882A37602
You can convert the STCK timestamp to a date and time using the LISTTOD command in IPCS:
IP LISTTOD B6CA230882A37602
QM1_2012-03-04_11.07.01
Product Synonym
IBM MQ WebSphere MQ WMQ
Was this topic helpful?
Document Information
Modified date:
21 June 2018
UID
swg21231008