Look at jboss's clustering framework.
If you write to this framework,
you can invoke methods on a cluster (either
synchronously or asynchronously)
Alternatively look at something like Jetty's session replication.
This uses an entity bean to replicate sessions.
Instead of using the JDBC persistence manager it
uses a persistent manager that replicates the state to
all nodes in the cluster.
Can I use this framework to get individual information on each node in a cluster? We want to do things like change the log level on and individual node or view the cache on an individual node.
To do this before we had map the logging object on each node in weblogic to a jndi name and access that name from our servlet for whatever node we wanted info on.
To do aggregation you would have to create an object that
is replication, similar to jetty's session store I mentioned before.
The object would have to include the node name in it
You can do the same in jboss as you do in weblogic using
Thanks! I have another question....Is there a method to get a list of all nodes in a cluster at runtime?