WLS Cluster debugging - some tips

Some tips when debugging WLS cluster stability

In one environment we had problems with cluster stability - and there were unnecessary migrations of JMS Servers.

Here are some settings that gave more info in logs. Set the following for the JVM:

-Dweblogic.debug.DebugServerMigration=true -Dweblogic.debug.DebugSingletonServices=true 

Change the below setting to DEBUG

 Home >Summary of Servers >AdminServer >Logging >General >Advanced >Minimum severity to log: 

enable below debugs from the weblogic admin console.

 Debug >weblogic >core 

cluster Collapse Node

 
 DebugAsyncQueue 
 DebugCluster 
 DebugClusterAnnouncements 
 DebugClusterFragments 
 DebugClusterHeartbeats 
 DebugLeaderElection 
 DebugReplication 
 DebugReplicationDetails 

Change the logging severity to Debug

 Servers-->Server_name-->Logging-->Advanced--> Standard out :Severity level:Debug 
 Log file :Severity level:Debug 

Restart.

Also look at Cluster » Monitoring » Summary to see the number of sent / received / retransmitted. For a healthy system sent and received should be almost same - and retransmissions low.