|
|
The WebSphere setup for this scenario is the same as for the Node Agent and appserver failover scenario described previously. The Deployment Manager has been installed on a remote system and the custom profile is installed into the HACMP cluster system. In addition, for this scenario, we installed an application that puts and gets messages to/from embedded messaging destinations.
For this scenario, make sure that the messaging engine (ME) runs only on the appserver that is running in the HACMP cluster. If you want to operate the ME in a specific appserver, you have to configure the proper core group policy for the ME. (For more information about core group policy configurations, see 6.2.4, Core group policy.) Therefore, we configured a static policy and relating match criteria as well as static group servers. We specified a certain messaging engine in the match criteria and selected the appserver running in the HACMP cluster as the static group server. Thus the messaging engine can now only run in our failover appserver.
