Clusters and Namelists

[BeanieNCecil]

In going over the IBM documentation on Clusters and Namelists I get so much information on these 5.2 features that I end up not really knowing for certain how these things will work.

1. Can a Namelist be constructed without using clusters?

2. If my only purpose is to make sure that two connected MQ Servers receive the same exact message packet why can't I just use clustering?

Thanks in advance for any help clearing this up.

 

[Guest_imported]

You can use clustering without any namelists for your purpose.

For example your queue managers are qmA and qmB and your queue name at both locations is inputQ, put these queue managers and and queues in a cluster lets say myCluster. It will work. namelists are useful when you have multiple clusters.

Thanks
Satish Basvapatri
satishb@hotmail.com

 

[smeridew]

Namelists and CLusters are two different things. A namelist just identifies more than one MQ object, and is usually used to reference more than one object at the same time (i.e. send to 2 queues at once). Clustering is used to reduce administration of channels, queues, etc. and also to load balance between multiple instances of queues.

If all you want to do is send the same message to two queues at the same time, just create a namelist with both queues identified.

If you want to send to one or the other queues (i.e. Q1 on QmgrA or Q1 on QmgrB) then use clustering. Scott Meridew
Certified MQSeries Specialist,MCSE
MQ Squared Inc.

http://www.mqsquared.com