To update spikes was common so we optimized a work flow.But wat could be the reason for spikes in single server.
From last error message I posted after optimising the fore said server is getting disconnected (or hanging ) and may be cluster is ping it...
Are you using the @Clustered annotation in your ejb bean .
Also you can set you loadbalancing policy as follows. This example shows the RoundRobin load balancing policy
I have 2 kinds of beans -Remote and local.My remote beans call the local beans.I have given clustered=true for remote beans and not local beans.Do you think this could be the reason.What is best take here?.My local beans are doing data intensive activities which take 2 to 3 mins and some time 5 mins.The load various a lot.
One more doubt .If I make the beans called from remote facade beans also remote and make them cluster will it get load balanced ?
What I mean is if I have remote bean A and B,C as local beans.Right now A is clustered.But if I make B and C remote and clustered will it get load balanced ?
If B and C are only called by A, and not by a client in an external process, no, making them remote/clustered will have no effect. JBoss detects the case where a call to a bean is coming from inside the same VM where the bean is deployed and optimizes the call to the in-VM bean rather than sending it remote.