Hi Hiram, Yep if you kill off the JVM of the broken publisher then things kick back into action....
Not sure this is good "expected behaviour". A misbehaving listener should not stop the publisher or other consumers from working. I would have hoped that the connections between the broker and the publishers and the consumers would be divorced so that this kinda of thing could not happen. Is there a way to configure ActiveMQ to do this ? Perhaps having seperate in-memory queues for consumers or by turning on persistance ? Jonathan On 3/7/07, Hiram Chirino <[EMAIL PROTECTED]> wrote:
if you kill the JVM of the broken listener does the publisher un-block? If so then this is the expected behaviour. The reason for this is that even though the listner is logically broken, (it's not receiving messages anymore), the broker does not know this since it still has an active connection with the broker. Furthermore, the broker is blocking the publisher because it is queueing up messages in memory for that dead consumer.
