Hi All,
We are successfully running a glassfish DAS with six running node agents and three clusters (two agents/instances per cluster). As part of failover testing for the DAS, we have replicated glassfish configuration data to another box on our internal network, and modified our DNS data to refer the node agents to the replicated DAS. To illustrate, here is our working configuration:
DAS box1 (gfadmin) --> agent1, agent2, agent3, agent4, agent5, agent6
Consider box1 the physical machine and gfadmin the DNS alias to the machine. All of the agents know the box through the alias gfadmin in their agent/config/das.properties file. When box1 is running the DAS, it sees all agents as running.
After bringing down the DAS on box1 and bringing up the DAS on box2, this is the new configuration:
DAS box2 (gfadmin) --> agent1, agent2, agent3, agent4, agent5, agent6
Note that DNS has been modified so that box2 is now referred to by the gfadmin alias. But when I log into gfadmin on box2, it does not see any of the agents as running. I tried restarting the agents as well as verifying that the agents and gfadmin on box2 are communicating over the network. Yet despite multiple restarts of the agents and the gfadmin on box2, it is unable to see the agents as running.
Does anyone who has expertise on the DAS/clustering/nodeagent subsystem on glassfish have any thoughts on what may be wrong?
Thanks,
-J
[Message sent by forum member 'jmetcalf' (jmetcalf)]
http://forums.java.net/jive/thread.jspa?messageID=264895