user1552907 wrote:Bug in the JNI code.
And for that we use JNI interface.
We did not observe any crashes when we changed the jvm to 1.5.0_11 from 1.6.0_27 version of JDK.Execution path changed - because the code did. So the bug you had before did cause the system to choke. Now it does.
user1552907 wrote:False cause.
Thanks for the reply.
We noticed that crash always happens at the time of allocating some object in jvm.
Even after disabling JVM option, "-UseBiasedLocking", we could see the crash, but the crash frequency reduced to a large extent. Now, we see that the crash occurs after the script ran for over 48 hours. Earlier, we could reproduce the crash in few minutes. Can we get any clue out of it?
Based on the above results, can there be possibility of issue lying in JVM (when it uses this option)?