Ungraceful publisher disconnects in HornetQ 2.2.8 EAP CR3

[bnc119]

Hi folks,

I am running hornet 2.2.8 EAP CR3.  I noticed something interesting in my hornetQ logs lately.  I have a client that connects to the server and creates creates 6 unique MessageProducers to 6 JMS topics .  They don't produce any messages.  When the client shuts down, it is seemingly going through the proper cleanup logic: i.e calling close() on it's connection object.

Yet after 30-60 seconds, the hornetQ log is flooded with the following sequence of warnings:

[hornetq-failure-check-thread] 1-Mar 17:41:44,99 WARNING [RemotingConnectionImpl]  Connection failure has been detected: Did not receive data from /<IP>:40337. It is likely the client has exited or crashed without closing its connection, or the network between the server and client has failed. You also might have configured connection-ttl and client-failure-check-period incorrectly. Please check user manual for more information. The connection will now be closed. [code=3]

* [hornetq-failure-check-thread] 1-Mar 17:41:44,99 WARNING [ServerSessionImpl]  Client connection failed, clearing up resources for session 87ce8055-63c5-11e1-9c6a-782bcb9e7c59

* [hornetq-failure-check-thread] 1-Mar 17:41:44,100 WARNING [ServerSessionImpl]  Cleared up resources for session 87ce8055-63c5-11e1-9c6a-782bcb9e7c59

I see this sequence 6 times, presumably one for each publisher.  Why doesn't HornetQ recognize that I've closed the connection?

How can I troubleshoot what is going on?  Jconsole only shows which subscribers are connected to a topic, not publishers.

Thanks!

[ataylor]

can you provide a simple test showing this

[bnc119]

Actually, I found my problem.  I was doing something dumb in my code.  I wasn't managing my publisher references correctly, and I accidently left some instances hanging around for HornetQ to find and purge for me, which it was doing correctly.