Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Won't Do
-
None
-
None
-
None
Description
Hi,
We are using in CERN ActiveMQ Apollo and before some months we used to have an "OutOfMemoryError". To limit this problem we decided to increase the Heap size allocated to JVM. In that spirit, we started running the broker in a fat box, featuring 48 GB RAM, and allocated 36 GB to the broker. This temporarily solve our problems, but from time to time our broker still gets stuck and we manually restart it.
However, now the logs say: "java.lang.OutOfMemoryError: unable to create new native thread".
After some reading, it seems that the more memory you give to the JVM heap the more likely you are to get "java.lang.OutOfMemoryError" unable to create new native thread exceptions when you have many threads.
Clearly, there is somewhere a leak and old threads are not removed. Consequently, our broker gets stuck after some days of healthy serving our consumers...Any plans of taking care of this bug?
Best Regards,
–
AV