| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| JMS to Clustered Queue only goes to local queue |
« View previous topic :: View next topic » |
| Author |
Message
|
| blegris |
 Posted: Fri Mar 30, 2007 6:16 am Post subject: JMS to Clustered Queue only goes to local queue |
 |
|
Newbie
Joined: 22 Mar 2007
Posts: 4
|
We have a JMS client (on WS Process Server) which writes a message to a clustered queue "Q1" which is defined on two queue managers QM1 and QM2.
I wanted to demonstrate that if I bring down a queue on QM1 that any messages sent to that queue on QM1 would make there way to QM2's Q1. To bring down the queue I would 'put inhibit' the queue. When I do this demonstration using the rfhutil tool it works as expected... the message sent to QM1 Q1 was sent from QM1 to QM2's Q1. However when we use the JMS client to send to QM1's Q1 thier is an error that indicates that the queue is ot available.
It looks like when I use RFHUTIL the clustering of the queue behaves normally, but when we use JMS to send the message the JMS client limited the cluster queue behaves as a single non clustered queue.
The JMS error is MQJMS2007: Cannot send msg to a queue with reason 2051
Any ideas on what could cause this?
Thanks for your help.
Brent |
|
| Back to top |
|
 |
| jsware |
| Posted: Fri Mar 30, 2007 7:26 am Post subject: |
|
|
Chevalier
Joined: 17 May 2001
Posts: 455
|
Are you using JNDI to bind JMS to your actual qmgr/queues? If so, ensure that the JMS queue object does not specify a qmgr name, otherwise you might be ending up with the quivalent of trying to put to a specific queue on a specific queue manager, bypassing queue resolution.
I am assuming you are using WMQ v6 and have correctly set the clustering options so that the queue manager does not favour a local instance of the queue over its clustered cousins all the time (since rfhutil works as you want).
_________________
Regards
John
The pain of low quaility far outlasts the joy of low price. |
|
| Back to top |
|
|
| blegris |
| Posted: Fri Mar 30, 2007 7:41 am Post subject: Resolved |
|
|
Newbie
Joined: 22 Mar 2007
Posts: 4
|
Thanks for your reply. We did the following to resolve the issue:
1) Our MQ server had the refresh pack 6.0.2 installed but our WPS was using the MQ classes that came with WPS, we changed the Environment variables in the WS admin console to point to the correct jar files. The variables that needed changed where MQJMS_LIB_ROOT and MQ_INSTALL_ROOT.
2) As you sugested we made sure that on the JMS Queue Destination we did not specify a queue manager name.
3) On the queue the CLWL use queue attribute (on the cluster page) was set to "queue manager" and on the Queue Manager the CLWL use queue attribute was set to 'ANY'. |
|
| Back to top |
|
|
| Michael Dag |
| Posted: Fri Mar 30, 2007 7:47 am Post subject: |
|
|
Jedi Knight
Joined: 13 Jun 2002
Posts: 2607
Location: The Netherlands (Amsterdam)
|
Good it's resolved!
in addition to:
| scottj2512 wrote: |
| I am assuming you are using WMQ v6 and have correctly set the clustering options so that the queue manager does not favour a local instance of the queue over its clustered cousins all the time (since rfhutil works as you want). |
I have actually seen this behaviour on V5.3.1 z/OS aswell, where I had QMA and QMB in a cluster and has CLUSQ on both Qmgrs.
WMB V6 was running on QMA and putting to CLUSQ on QMA, when I put inhibited CLUSQ on QMA, the messages went automatically to CLUSQ on QMB ...
_________________
Michael
MQSystems Facebook page |
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
