Art, Response is inline. > Hmm, what you describe sounds disturbing and counter to my experience. > > Opening a jira entry is the first step to getting any bug addressed. > Attaching a self-contained junit that can reproduce the problem certainly > helps a lot.
I'll ask to my society if I can allocate some time for this task. > One question comes to mind - is a connection pool part of the setup? If so, > remember that calling close on the connection returned by the pool does not > necessarily close the underlying connection (otherwise, it wouldn't be much > of a pool). So, in that case, make sure the pool is being shutdown > properly; it should then make sure the connection is shutdown. Yes, I'm using a connection pool instantiated directly into the code (no jndi) and I'm calling clear() (which should close all the connections) then close() over it. This is done into a spring dispose() call (so this code is actually executed at the context dispose). > Again, though, if you can prove a case that closing the connection leaves > the transport open, please create a Jira entry - we'll need to get it > solved. I'll try to give u a junit test if it will be possible anyhow looking at the method I pointed u in the last mail it should be easy to understand thst the Transport is never closed since (as far as i understood) the pool is able ro close the connections but no one is taking care of the Transport. Cheers, Carlo > > > > -- > View this message in context: http://activemq.2283324.n4.nabble.com/5-8-0-Unable-to-shutdown-embedded-broker-tp4679137p4679141.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com.
