In my Application when running in production server it get hanged(that hang is due to toplink hang after i analyzed) .No Deadlock is found.And from application point of view there is no locking in that particular functionality.Please suggest some possible scenario that could make toplink hang.Below are the two thread dumps.
Application Architecture is
JSF & ADF(View)-Bean-Delegator-EJB Session Bean-Facade
Production Server- Oracle Application Server
From Thread Dump no deadlock is there,I am able to find some of the methods from my application level class except framework class from the thread exception stack trace but i am unable to detect the root cause taht means which type of code(possible causes that can lead to toplink related problem) can cause the toplink hang.Please suggest some advice.
Please see http://wiki.eclipse.org/EclipseLink/FAQ/JPA#How_to_diagnose_and_resolve_hangs_and_deadlocks.3F for info on diagnosing deadlocks and hangs within TopLink/EclipseLink. If that does not help, I'd recommend you file a support case to go over the thread dump and sequence of events that lead up to the problem.
Thanks for the suggestion. i go through the link you have included in your thread.Actually the method i traced is a very big method(2000 lines approx) and from this it is calling several big methods which are also using unitofwork of toplink including the original method .So is it causing to hang toplink?
The thread dump data is actually showing symptom of heavy JVM contention. When you see this pattern, this normally indicates either excessive CPU/IO utilization from the OS/host or excessive garbage collection from the JVM.So in previous thread you suggested to file a support case to go over the thread dump and sequence of events that lead up to the problem.So can you give me any specific solution to resolve this issue.
Sorry, you have not provided enough information to make suggestions. You mention a thread dump, but I don't see it attached to the thread which is partly why I suggest using support - that, and this will require application tuning which is difficult to suggest without having the big picture of how the application works, its load and usage paterns etc.
Actually the thread dump file was too long(more than 30000 character ) thats why it was not allowing me to submit after pasting the content of that file.
i analyzed the thread dump, it is a case of heavy Thread contention which is leading to high CPU usage eventually causing hang in production server.
But i am unable analyse/get the root cause and also not able to find any application specific solution.
It will be better if you post your mail id or you can send it to my mail id : firstname.lastname@example.org so that i will be able to send those thread dump file to you.
Actually the thread dump file is very long which is exceeding the limit so that it is not allowing me to post here.So can u send your email-id to my id email@example.com so that i will be able to send the thread dump file.