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 » Cluster GET question

Post new topic  Reply to topic
 Cluster GET question « View previous topic :: View next topic » 
Author Message
paul52
PostPosted: Sun Aug 18, 2002 6:25 am    Post subject: Cluster GET question Reply with quote

Novice

Joined: 01 Jul 2002
Posts: 22

Hello Everyone

Thank you in advance for help. Here is my cluster test scenario.

I have 3 Queue managers in Cluster network with the following cluster queues.

QM1: CLQ_01, CLQ_03, CL_05

QM2: CLQ_02, CLQ_04, CL_06

QM3: CLQ_02,




1. Both Application A & Application B are connected to QM2. Let's say
An application connected to QM1 issue puts 5 messages(amqsput CLQ_2 QM1 five times) to cluster queue CLQ_02 hosted by QM2 & QM3. Now there are 5 message in CLQ_02. I noticed Message 1, 3, 5 ends up CLQ_02 in Queue Manager QM2 and Message 2, 4 ends up
in CLQ_02 in QM3. I guess this is expected because of workload balancing. I understand message will be distribued in round roubind fashion and message migh end of any of the cluster queue

Now my guestion is:
How will Cluster GET works in workload balancing. I understand you can only GET from local QM.. Let's say, If i issue MQGETBR (Bowse message)
from QM2, I can only see message 1,3,5 because message 2, 4 are in CL_02 in QM3. How my application
connected to QM2 will get message 2, 4. Since Message 2 & 4 are in CLQ_02 in QM3 and i can NOT issue MEGET to message 2, 4.
How MQSeries Clustering handle this type of situation.


2. What would happen the messages hosted by CLQ_02 in QM3 (e.g Message 2 & 4 in my case) if QM3 is down or crashed.

Thanks again
Back to top
View user's profile Send private message
bduncan
PostPosted: Sun Aug 18, 2002 5:26 pm    Post subject: Reply with quote

Padawan

Joined: 11 Apr 2001
Posts: 1554
Location: Silicon Valley

1. Yes, you can only retrieve messages from queues on the same physical machine as your application is running on. In this case, the application on QM2 will only retrieve messages 1,3, and 5. The only way to get messages 2 and 4 is to have another application running on QM3. This is how clustering works. When you setup your cluster, the fact that you put two instances of CLQ_02 on QM2 and QM3 means that you don't care which instance of CLQ_02 a message goes to. That means that whatever processing which occurs after the message is retreived, should be indentical on QM2 and QM3. So this means you need multiple copies of your application running on both machines. Don't think of this as a drawback or disadvantage; it's simply how a cluster works. It sort of defeats the purpose of sending the message over the network to QM3 in the first place if your MQGET meant that it had to travel over the network back to QM2. Which begs the question of why you created CLQ_02 on QM3 in the first place.

2. What happens to those messages on QM3 (message 2 and 4) depends on whether they are created as persistent messages. If they are non-persistent, then when QM3 comes back online, they will be gone. If they are persistent, then while QM3 is down, they will be unavailable for MQGET, but when QM3 comes back online, an application running on the same machine will be able to retrieve the messages.
_________________
Brandon Duncan
IBM Certified MQSeries Specialist
MQSeries.net forum moderator
Back to top
View user's profile Send private message Visit poster's website AIM Address
jc_squire
PostPosted: Sun Aug 18, 2002 6:48 pm    Post subject: Reply with quote

Centurion

Joined: 14 Apr 2002
Posts: 105
Location: New Zealand

Hi Paul,

FYI - you do not have to apply work load balancing. If you look at the IBM MQ manual on clustering it will explain how to apply a workload exit (supplied with the product code) to route messages to a specific queue manager and if that queue manager is not available the messages can be routed to a second queue manager (an instance of the same queue exists on both queue managers). You can also write your own workload exit, sample code etc in the manual.

Regards
_________________
J C Squire
IBM Certified Specialist - MQSeries
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 » Cluster GET question
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.