I have to hang my head in shame, it seems that the issue is not related to
the engine, but a linked blocking queue that I was using to use to
communicate with another thread. My hypothesis is that with the "update"
approach, the first thread is going too fast and the second thread simply
cannot keep up, where as with the insert approach, the first thread is
slowed down enough to allow the second thread to keep up.
The heapdump analysis showed that it was only the queue that was keeping the
references.
I will delete this post.
--
View this message in context:
http://drools.46999.n3.nabble.com/Possible-memory-leak-in-5-3-with-update...
Sent from the Drools: User forum mailing list archive at
Nabble.com.