| Author |
Message
|
| narendra68 |
 Posted: Fri Apr 15, 2005 7:03 am Post subject: Request times out bcaz Cluster chls are inactive |
 |
|
Apprentice
Joined: 17 Nov 2004
Posts: 30
|
Gurus,
We have 5 qmgrs in cluster. 4 on NT and 1 on AS400.
We have 2 message flows running on two qmgrs .
First message flow takes request and routes to AS400.
Second one process the reply from AS400 client.
Everythinng works when all the channels are running.
THE PROBLEM IS THE FIRST HIT IN THE MORNING WHEN ALL THE CHLS ARE IN INACTIVE STATE, THE FIRST REQUEST TIMES.
WE HAVE THE APPLICATION WAIT INTERVAL AET TO 12 SECONDS.
Would some please tell me what is the best way to solve this problem
Thanks
Narendra 
_________________
Narendra |
|
| Back to top |
|
 |
| PGoodhart |
| Posted: Mon Apr 18, 2005 10:00 am Post subject: |
|
|
Master
Joined: 17 Jun 2004
Posts: 278
Location: Harrisburg PA
|
Wait longer?
Seriously does it come back without action? Or do you have to monkey the channels...
_________________
Patrick Goodhart
MQ Admin/Web Developer/Consultant
WebSphere Application Server Admin |
|
| Back to top |
|
|
| narendra68 |
| Posted: Mon Apr 18, 2005 10:50 am Post subject: |
|
|
Apprentice
Joined: 17 Nov 2004
Posts: 30
|
PGodhart
We need wiat very long or we need to start the channesl.
Thanks
_________________
Narendra |
|
| Back to top |
|
|
| EddieA |
| Posted: Mon Apr 18, 2005 12:56 pm Post subject: |
|
|
Jedi
Joined: 28 Jun 2001
Posts: 2453
Location: Los Angeles
|
| Quote: |
| THE PROBLEM IS THE FIRST HIT IN THE MORNING WHEN ALL THE CHLS ARE IN INACTIVE STATE |
Are you certain that it's a MQ problem. Could it be that the first time, each day, that the AS400 application is called, it takes a long time to be initiated and reply.
Cheers,
_________________
Eddie Atherton
IBM Certified Solution Developer - WebSphere Message Broker V6.1
IBM Certified Solution Developer - WebSphere Message Broker V7.0 |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Mon Apr 18, 2005 2:32 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
We have an app that needs a reply in <2000 ms, and it includes many hops through a cluster including WB-IMB work. Triggering of all these channels was enough of a delay to cause the 1st message of the day to time out.
So we made the DISCINT of the CLUSRCVRs 300,000. This is long enough to keep the channels running through a long weekend of no activity, so Monday's 1st message did not time out.
But you say you are waiting 12,000 milliseconds, There is no way MQ channels would trigger that slowly. You have another problem, most likely what Eddie said, the app needs to inititalize itself on the first message every morning and it si to slow, not MQ.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| PGoodhart |
| Posted: Tue Apr 19, 2005 3:26 am Post subject: |
|
|
Master
Joined: 17 Jun 2004
Posts: 278
Location: Harrisburg PA
|
Yes, it sounds like your application is going through some serious "wake-up" routine. What is the application written in? Java? Do you have megabytes of configuration files to parse or something maybe?
_________________
Patrick Goodhart
MQ Admin/Web Developer/Consultant
WebSphere Application Server Admin |
|
| Back to top |
|
|
|
|
