OML9.10 RHEL Cluster - nodes buffer message when cluster is switched
I have a weired issue going on.
we have OML9.10 on RHEL 5.6 cluster.
We have ov-server package on NODE1 and ov-oracle package on NODE2.
The application runs fine as per above set-up. When we move the ov-server package to NODE2, the application is up and runs fine but the managed nodes are buffering messages, no messages arrive to management server.
Once I switch back the ov-server package from NODE2 to NODE1 everything runs fine, all the messages from the node are seen.
has anyone come acroos such a kind of weired issue, if yes please help me in getting the issue resolved. We are about to move the server to production in a week and just found out this issue.
Thanks in advance,
Have a look at the document High Availability Through Server Pooling on to check if OML and your agents are correctly configured.
Does the nodename referenced in the buffering reference the OM Package hostname?
Does this name resolve to the relocatable package IP Address?
Check your nodes and make sure they expect the Management Server to be on the relocatable address associated with the package.
If they are looking for the Management Server on Node1 then they will need to be reconfigured to associate the Management server with the package address.