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 » Clustering » Basic Clustering Concepts

Post new topic  Reply to topic
 Basic Clustering Concepts « View previous topic :: View next topic » 
Author Message
JohnRodey
PostPosted: Tue Jun 28, 2005 12:00 pm    Post subject: Basic Clustering Concepts Reply with quote

Centurion

Joined: 13 Apr 2005
Posts: 103
We have just recently started looking into clustering queue managers.

A coworker of mine was under the impression that you can somehow simply put a message to a certain queue, and not specify a QM, and that message will be routed to the correct QM. Sort of like saying... "imqsput MYQUEUE" and somehow your cluster will determine which QM's have a "MYQUEUE" and round robin it to one of those QMs.

(Even if this was capable I do know that "imqsput MYQUEUE" would just use your default QM)

My idea of what clustering is used for is that you would want to place a message on a full repository (or possibly partial repository) queue manager then it would be routed (round robin) to a QM that had that queue in the cluster.

Now if the machine with the full repository were down we would have to then write that message to another full repository QM, correct?

I guess I'm just looking for assurance that I am thinking about things right and whether or not my coworkers concept is a reality?

Thanks in advance
Back to top
View user's profile Send private message
fjb_saper
PostPosted: Tue Jun 28, 2005 12:45 pm    Post subject: Reply with quote

Grand High Poobah

Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
Not quite.

In order to get the message "round robin" you do not want the queue to exist on your "gateway". The gateway will then find it in the cluster and deliver the message to it.

If the qmgr info is blank on the destination qmgr field (can be achieved through a remote qdef or qmgr/cluster alias) then the qmgr looks to the cluster for resolution.

Read the cluster manual for more detail. And no the gateway does not need to be a full repository but it is nice if it is as it allows global visibility of the cluster. And Yes you should have at least one full repository alive and active in the cluster at all times. This is why it is prudent to always have 2 full repositories in the cluster at all times.



Enjoy
Back to top
View user's profile Send private message Send e-mail
PeterPotkay
PostPosted: Tue Jun 28, 2005 5:20 pm    Post subject: Reply with quote

Poobah

Joined: 15 May 2001
Posts: 7722
fjb_saper wrote:
In order to get the message "round robin" you do not want the queue to exist on your "gateway". The gateway will then find it in the cluster and deliver the message to it.

In the current version of MQ, you now have the option of continuing the round robin behaviour to all QMs in the cluster that host your q, even if the local connected-to QM has that queue as well. This is new in 6.0. fjb_saper's answer is correct for all MQ version prior to 6.0.
_________________
Peter Potkay
Keep Calm and MQ On
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 » Clustering » Basic Clustering Concepts
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.