[jboss-user] [JBoss Messaging] - Re: PingTimerTask broken pipe

[bob_walker99]

Hi - I' checked my code and did find some instances in my sending client that wasn't closing connections properly in the event of exceptions, which I've fixed, but I've had the same problem overnight again.

I've isolated my sending and receiving jobs - this morning I've successfully sent around 10,000 messages to a single queue with no apparent resource hit to either my client or the JBM server, so I have to conclude that the problem is in the receiver. 

I have an ExceptionListener in place that notifies the thread that starts the listener when an exception occurs, this disposes all the JMS environment objects and re-initialises them, correctly as far as I can see.

However, all this is kind of academic, because I'm not actually getting exceptions in my listener code - the threads and open file handles are just mounting and mounting, but until the server runs out of resources, it does actually work. I'm at a bit of a loss as to where to look next...?

If it helps, these are fairly large text messages, often up to 1Mb, so I've set the queue's FullSize param to 100, and the PageSize and DownCacheSize to 20 apiece -  are these reasonable values?

Any advice gratefully received.

Best regards,
Bob

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4060400#4060400

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4060400