ASG
IBM
Zystems
Cressida
Icon
Netflexity
 
  MQSeries.net
Search  Search       Tech Exchange      Education      Certifications      Library      Info Center      SupportPacs      LinkedIn  Search  Search                                                                   FAQ  FAQ   Usergroups  Usergroups
 
Register  ::  Log in Log in to check your private messages
 
RSS Feed - WebSphere MQ Support RSS Feed - Message Broker Support

MQSeries.net Forum Index » Clustering » Cluster workload exit rejecting messages.

Post new topic  Reply to topic
 Cluster workload exit rejecting messages. « View previous topic :: View next topic » 
Author Message
aboggis
PostPosted: Mon Feb 09, 2004 3:02 pm    Post subject: Cluster workload exit rejecting messages. Reply with quote

Centurion

Joined: 18 Dec 2001
Posts: 105
Location: Auburn, California

I have a Solaris based environment (5.8, all running WMQ 5.3 CSD05) with 6 queue managers involved across four hosts.

All the queue managers are running.

The issue I am having is that periodically the cluster workload exit is rejecting messages that are put to a remote cluster queue when it thinks that the channel to the remote manager is not up. However, there is no corresponding channel event logged (on either the sending or the receiving queue manager). We need to have this exit in place since we require that the putting application receives an error if the intended message destination is unavailable. This way we can implement failover processing for our time-critical messages.

The exit is basically a simplified version of the sample exit provided (amqswlm0.c).

As a work around we defined an addition CLUSSDR (locally) and a CLUSCVR (on the destination) and our messages started flowing again.

Can anyone provide any insight as to why the exit might reject messages even though no channel events were logged?

Regards,

tonyB.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
aboggis
PostPosted: Tue Feb 10, 2004 12:03 pm    Post subject: Reply with quote

Centurion

Joined: 18 Dec 2001
Posts: 105
Location: Auburn, California

More information... We've added some logging to our exit and are seeing some interesting results:

Code:
MQWXP * parms

This is passed to the exit by MQ.

If parms->ExitReason == MQXR_INIT, we return parms->ExitResponse2 = MQXR2_DYNAMIC_CACHE to indicate that we supprt dynamic caching.

If parms->ExitReason != MQXR_CLWL_PUT, we return, having previously set parms->ExitResponse = MQXCC_OK, since we only care about controlling MQPUT's.

So, the very next things we do is check the target queue manager's availability:

Code:
qmgr = parms->DestinationArrayPtr[ 0 ]
if ((qmgr->QMgrFlags & MQQMF_AVAILABLE) == 0)
{
parms->ExitResponse = MQXCC_SUPPRESS_FUNCTION;
return;
}

This is causing the exit to exit (!). However, I *KNOW* the remote queue manager is available because I am connected to it, locally using runmqsc and via Websphere MQ Explorer. If I disable my exit, I can put successfully too. The value in qmgr->QMgrFlags is 0x225B, which (if my math is correct) equates to: (MQQMF_REPOSITORY_Q_MGR & MQQMF_CLUSSDR_USER_DEFINED & MQQMF_CLUSSDR_AUTO_DEFINED), plus a bunch of other flags (which the docs say are used internally). What I don't understand is how both USER... & AUTO... are set.

tonyB.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Display posts from previous:   
Post new topic  Reply to topic Page 1 of 1

MQSeries.net Forum Index » Clustering » Cluster workload exit rejecting messages.
Jump to:  



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
Protected by Anti-Spam ACP
 
 


Theme by Dustin Baccetti
Powered by phpBB © 2001, 2002 phpBB Group

Copyright © MQSeries.net. All rights reserved.