Thank you for you reply, ceposta.
ceposta wrote > The broker has fired an advisory because a new consumer has been added. > But > you're contended around the lock to send messages because some other > thread > has the lock (dead lock possibly)? >From this thread trace, we can tell that the message couldn't be sent out. What would happen then? Assuming the consumer was the one that was going to consume the messages in the target queue that this advisory was aimed to add into, would it be available to consume the messages locally or not? As the stack trace shown, it has a one way dependency up to the notFull condition. It looked like thread starvation instead of dead lock. If the consumer was available even without sending out the advisory message successfully, then this was not our culprit. But if the consumer wouldn't be available until the advisory message was sent out successfully, this looked like the cause of the thread starvation. What do you think? Thanks, Jeff -- View this message in context: http://activemq.2283324.n4.nabble.com/ActiveMQ-5-5-1-Communication-got-stuck-between-Producer-and-Consumer-tp4671812p4671832.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
