I am hunting for a quick solution to the problem that we are facing. Let me just make you understand the problem in a very simple manner.

We have an enterprise java web-application in which the entire architecture of the application is maintained under 4 distributed servers. When we deploy the code artifacts(binaries/jar/ear files) in 2 different servers, both of them behave differently. When I say server I mean  JBOSS (business logic) server. What I mean to say is, when the binaries is deployed in 1 of the server, all work as expected. But when deployed on the other similar server, occasionally the memory consumption by the JAVA service goes very high and reaches the threshold limit and then the application behaves erratically / responds very slowly / hangs / gives 502 server error.

So, my blind guess to this issue is going towards the JBOSS server configuration. To the best of my knowledge, the JBOSS server configuration for both the above said servers (in which we deployed our artifacts) are same.

Does any of you guys, have any specific idea about the JBOSS server configuration parameter, which might have potential impact to cause this issue that I am facing ?

All kind of suggestions / experiences sharing are most welcome. Waiting in anticipation to your valuable inputs.