To enable high availability and workload management in the cluster, each messaging engine needs a core group policy. These policies control the distribution of the messaging engines, defining the preferred cluster member to use. If that cluster member fails, the messaging engine switches to another cluster member, but returns to the preferred cluster member when it becomes available.
Perform the following procedure once for each cluster member. The procedure provides the required steps to create a core group policy. For more information on this topic, see Setting up Preferred Servers in the Default Messaging Provider section of the WebSphere documentation.
To create a core group policy
Note: Do not delete or modify the default policies.
WebSphere automatically generates the name of the messaging engine when you create it. The name has the following format:
cluster_name.00n-IMSBus
where cluster_name is the name of the cluster you are configuring, and n represents a unique number for the messaging engine, which is automatically incremented each time a messaging engine is created for the cluster.
For example, if the cluster name is im_cluster, and there are two messaging engines, the names would be:
im_cluster.001-IMSBus
im_cluster.002-IMSBus
The message engine belongs to the cluster member when the JNDI name contains the cluster member’s node name.
You can select as many cluster members as needed from the cluster where the messaging engine is defined. Do not select node agents or the Deployment Manager.
The first cluster member in the list is the one that the messaging engine will use by default. Move the cluster member up or down in the list until they appear in the order in which they should be used.
| Copyright © 2010 CA. All rights reserved. | Email CA Technologies about this topic |