| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Clustered QM goes down |
« View previous topic :: View next topic » |
| Author |
Message
|
| PeterPotkay |
 Posted: Tue Nov 08, 2005 6:26 am Post subject: Clustered QM goes down |
 |
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
I'm gonna build a little cluster to test this, but would appreciate some "official" insight. Couldn't find anything in the manuals that specifically covered this.
QM1 and QM2 are PRs.
QM3 and QM4 are FRs.
QueuePeter is hosted on QM1 only, and is clustered.
App on QM2 is happily putting to QueuePeter via the magic of clustering.
QM1 goes down.
What happens to the app on QM2 when it does the next MQPUT1 to QueuePeter? Does it queue up in SYSTEM.CLUSTER.TRANSMIT.QUEUE on QM2 until QM1 becomes available. Or does the MQPUT1 fail, because at that moment there is no longer a valid target?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Tue Nov 08, 2005 6:41 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
I think qclusters are treated as any other qremote from the point of view of MQPUT.
But you might get a 2269 or 2270.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Tue Nov 08, 2005 7:04 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
OK, I tested this out. Once I took QM1 down, the app on QM2 (amqsput) was still able to keep putting, wheter it was additional puts over an already open q, or a new session. The messages just queued up in the SYSTEM.CLUSTER.TRANSMIT.QUEUE. Once QM1 came back up, all the messages arrived.
If the messages were persistent, not a single one was lost.
If the messages were non persistent, the first one after QM1 goes down is ***gone***. This was over a FAST channel.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| wschutz |
| Posted: Tue Nov 08, 2005 7:11 am Post subject: |
|
|
Jedi Knight
Joined: 02 Jun 2005
Posts: 3316
Location: IBM (retired)
|
Peter, I think that behaviour is documented in the cluster workload algorithm, step ( ie "If less than one remote instance of a queue remains, all MQCHS_RETRYING channels are selected" and step (9) REQUESTING, PAUSED, STEPPED channels.
_________________
-wayne |
|
| 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
|
|
|
|
