| Author |
Message
|
| Sam Uppu |
 Posted: Thu Oct 28, 2010 4:33 am Post subject: Input queue filledup/NO IPPROCS and flow not processing msgs |
 |
|
Yatiri
Joined: 11 Nov 2008
Posts: 610
|
Hi,
We are doing a stress test for our flows. One of the flow is not pulling the msgs from input queue though the flow/ execution group are up and running. The input queue is filled up and when we look at the input processes(IPPROCS), it is zero. Not sure why the dataflowengine is detached from the queue. We are using nonpersistent msgs.
Can you guys give some scenarios where the dataflowengine may detach from input queue and msgs stack up input queue?.
thanks |
|
| Back to top |
|
 |
| Vitor |
| Posted: Thu Oct 28, 2010 4:57 am Post subject: Re: Input queue filledup/NO IPPROCS and flow not processing |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| Sam Uppu wrote: |
| Can you guys give some scenarios where the dataflowengine may detach from input queue and msgs stack up input queue?. |
When it crashes? Anything good in the Event viewer / logs?
How many instances of the flow are there? If only 1, have you tried multiple instances?
What version of WMB? What platform?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| Sam Uppu |
| Posted: Thu Oct 28, 2010 6:14 am Post subject: Re: Input queue filledup/NO IPPROCS and flow not processing |
|
|
Yatiri
Joined: 11 Nov 2008
Posts: 610
|
| Vitor wrote: |
| Sam Uppu wrote: |
| Can you guys give some scenarios where the dataflowengine may detach from input queue and msgs stack up input queue?. |
When it crashes? Anything good in the Event viewer / logs?
How many instances of the flow are there? If only 1, have you tried multiple instances?
What version of WMB? What platform? |
Sorry..Vitor,
Here is the info:
We are on AIX 6.1, WMB 7.0.0.1, MQ 7.0.1.3
We are using 10 additional instances on the flow.
I dont see good info in the broker syslogs when I glanced but I need to look closely into the broker syslogs.
No abends, no cores, no QMgr FDCs and no restart of the execution group seen.
Let me know your thoughts.
Thanks, |
|
| Back to top |
|
|
| inder |
| Posted: Thu Oct 28, 2010 6:14 am Post subject: |
|
|
Apprentice
Joined: 24 Mar 2003
Posts: 49
Location: USA
|
It normally happens when you are using message grouping. and got to do with sequencing especially in testing environments.
If you specify what the flow is doing ang what type of messages are being processed, we can throw some light
regards
Inder |
|
| Back to top |
|
|
| zonko |
| Posted: Thu Oct 28, 2010 7:17 am Post subject: |
|
|
Voyager
Joined: 04 Nov 2009
Posts: 78
|
Back to basics:
Maybe the queue name and/or qmgr name on the flow do not match the real queue and qmgr names. |
|
| Back to top |
|
|
| Sam Uppu |
| Posted: Thu Oct 28, 2010 8:08 am Post subject: |
|
|
Yatiri
Joined: 11 Nov 2008
Posts: 610
|
| inder wrote: |
It normally happens when you are using message grouping. and got to do with sequencing especially in testing environments.
If you specify what the flow is doing ang what type of messages are being processed, we can throw some light
regards
Inder |
We are not using any message sequencing. Those are small messages with 25kbyte without dependency on each other. |
|
| Back to top |
|
|
| Sam Uppu |
| Posted: Thu Oct 28, 2010 8:13 am Post subject: |
|
|
Yatiri
Joined: 11 Nov 2008
Posts: 610
|
| zonko wrote: |
Back to basics:
Maybe the queue name and/or qmgr name on the flow do not match the real queue and qmgr names. |
We processesd the messages with 20 / second for 4 hours and we see this issue. If we have any issue with the queues/ Qmgr names, we could have breakdown before 4 hours itself and not running for 4 hours successfully.
Thanks, |
|
| Back to top |
|
|
|
|
