| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Multiple Execution groups/Multiple message flows |
« View previous topic :: View next topic » |
| Author |
Message
|
| venkat kurra |
 Posted: Thu Oct 16, 2003 11:45 am Post subject: Multiple Execution groups/Multiple message flows |
 |
|
Master
Joined: 18 Oct 2001
Posts: 245
Location: Bloomington , IL
|
Hi,
Any body explain how the execution group or message flow uses CPU time and memory?
An application put thousends of messages on WMQI Input queue and that queue is backing up for some time. System has 2 CPU's with 2GB(main memory). Shell i create one more execution group with same message flow or deploy one more message flow with same nodes(same flow with different name).Which one is the best? are these execution groups /message flows run as saparate threads or processes?
_________________
Thanks,
Venkat Kurra
IBM Certified Specialist-MQSeries
IBM Websphere MQSeries Administrator
IBM WebSphere Message Broker System Admin |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Oct 16, 2003 12:57 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
First, add more 'additional instances' to your already deployed flow in your Operations tab. This will allocate more threads within the execution group process to your flow. You will see the 'open input count' on your queue go up.
Then, to take better advantange of both CPUs, create an additional execution group (a new process), and deploy the same flow to it.
A copy of the message flow with a different name assigned to the same execution group is the equivalent to adding one additional instance. But it's messier, and harder to maintain. So don't do it that way, just add more instances.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| wmqiguy |
| Posted: Fri Oct 17, 2003 4:52 am Post subject: |
|
|
Centurion
Joined: 09 Oct 2002
Posts: 145
Location: Florida
|
One more quick note on all of this. Make sure that you do not have any affinities between the messages. If the processing of the second message is dependent upon results of the first message, then you are stuck with "no additional instances". If you don't have any affinities, congratulations, you have fairly smooth sailing. (Why do I always end up with interfaces with affinities...was I bad in a previous life?!?!?!)
Another thing about a lack of affinities, is that it gives you the ability to throw money at the problem with more hardware (and licensing). IF this is the case, think about creating a broker on multiple machines and use clustering to load balance.
Have fun!  |
|
| 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
|
|
|
|
