| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Message Flows seem to get stuck in EG on WBIMB v5 CSD04 |
« View previous topic :: View next topic » |
| Author |
Message
|
| williad |
 Posted: Thu Sep 08, 2005 1:21 am Post subject: Message Flows seem to get stuck in EG on WBIMB v5 CSD04 |
 |
|
Apprentice
Joined: 08 Sep 2005
Posts: 40
Location: London
|
HI
We have WBIMB v5 CSD04 on a sun solaris v9 server.
The problem we are experiancing is explained as best below:
A particular execution group seems to get stuck whereby all the message flows stop reading from their input queues, hence the sending applictaion to these queues fill up the queues to their max depth. The strange thing is that the open input count for these queues seems to imply that the message flows are still attached to the queues.
There are no errors that i can relating to the above problem in the error log on the server. When i do a mqsi <broker> -e <execution group> all seems to be ok. The number of dataflowengine processes match the number of executions groups on the broker.
I hope i explained this ok
any help/suggestions will be greatly appreciated |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Sep 08, 2005 4:03 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
Check the retry count on the first message in the queue.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| mq_developer |
| Posted: Thu Sep 08, 2005 10:24 am Post subject: |
|
|
Voyager
Joined: 18 Feb 2002
Posts: 82
|
When you see execution group / broker is not working as you expected it to ... this is what i would normally do
1) Issue DIS QSTATUS(<MessageFlowInputQueue>) TYPE(HANDLE) OPENTYPE(INPUT) ALL
Get the process id from PID attribute
2) Issue mqsilist <BrokerName>
Check that process id returned from step 1 matches with that off the execution group which you think is actually running the flow
3) If everything looks good so far ,
Get disable the <MessageFlowInputQueue> , if the execution group is repsonding , you would MQGET failure with reason code 2016 in your syslog
4) Still everything seems fine , browse your message and look at the backout thershold , if it keeps increasing ... Your message flow is in retry loop because of inadequate error handling, its the message flow which is culprit.
5) once again its all good , look at the CPU usage on your broker server (command differs from OS to OS) ..
If you see your execution group at the top hogging CPU / Memory , it means your message flow is working , but its on a Endless loop somewhere within or waiting on some resouce ..
6) Finally nothing seems a problem , Call IBM suppport 
Hope this helps,
Ram. |
|
| Back to top |
|
|
| williad |
| Posted: Fri Sep 23, 2005 7:09 am Post subject: |
|
|
Apprentice
Joined: 08 Sep 2005
Posts: 40
Location: London
|
Hi
thanks for all your responses. the flow was in a loop.
the message flow had additional instances
and it seems the logic of the flow was to write to a database table
and then when the next msg comes in look at that table for a value. It seems like the msg was loopoing looking for that value before the value was inserted into the table
thank you all again
williad |
|
| 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
|
|
|
|
