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 » WebSphere Message Broker (ACE) Support » Finding my performance constraint

Post new topic  Reply to topic
 Finding my performance constraint « View previous topic :: View next topic » 
Author Message
goffinf
PostPosted: Tue May 12, 2009 4:12 pm    Post subject: Finding my performance constraint Reply with quote

Chevalier

Joined: 05 Nov 2005
Posts: 401
Broker 6.1
AIX 5.3
4GB RAM
CPUs allocated on demand

Sorry for the very general nature of this question, but I'm just looking for the most likely causes that we should investigate.

We have a flow which receives a message and makes a 5 external calls before returning. The calls have dependencies so have be executed one after the other. The processing is CPU bound. Under light load everything is fine, the CPUs 'light up' to a reasonably high percentage. However when we ramp up the concurrency beyond a certain point the %CPU usage goes down !! , there is some constraint to thruput here that we can't figure out. The performance of the external calls remains high.

We increased additional instances of the flow and initially saw roughly linear scaling. However once we increased additonal instances of the flow to cope with the maximum arrival rate of messages that we will need (I think we set it around 50) we see a dramatic fall off in thruput, i.e the end-2-end performance degrades (a lot).

Is it likely that resources required by Broker to support the additonal instances are running short ??

Fraser.
Back to top
View user's profile Send private message
recallsunny
PostPosted: Wed May 13, 2009 12:30 pm    Post subject: Reply with quote

Disciple

Joined: 15 Jun 2005
Posts: 163
Location: Massachusetts
50 instances (per flow) and each one operating 5 threads (for responses) puts this right at the operation system limit for the no.of practical concurrent threads per EG. IBM Recommended working limits are 256 threads per process for Windows and UNIX and 230 TCBs per Address Space on z/OS

The best way to mitigate this would be to spread the instances of the flow to different execution groups. Hope this helps!
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic  Reply to topic Page 1 of 1

MQSeries.net Forum Index » WebSphere Message Broker (ACE) Support » Finding my performance constraint
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.