You say you're doing request-response with a 60-second timeout on the response. If some of those responses took longer than 60 seconds, I'd expect that that particular response message wouldn't be removed from the response queue, and after some number of them I'd expect that the response queue would be full of messages that were never going to be consumed and producer flow control would kick in for the response queue. Depending on how you've implemented your producer, this might or might not prevent it from publishing any more requests. At a minimum, you should make sure you've thought about what should happen to messages that are not consumed due to the 60-second timeout, even if this doesn't turn out to be the root cause for your problem...
Do you have any messages in the broker logs about producer flow control? And do you have any way to find out how many messages were in the response queue at the time this all went down? Also, are you sure you're closing all of your consumers properly? Selectors are expensive for a broker to evaluate, so if it's still evaluating whether to deliver each message to every single consumer that's ever connected, that might explain the slowdown you see (though it doesn't explain why it completely stops processing data, so this may not be it). On Mon, Jan 12, 2015 at 4:03 AM, xabhi <xabh...@gmail.com> wrote: > Please help i am not able to get anywhere with this issue. > > Atleast tell me what more information to collect and what things to look > out > for next time this issue happens which can help to resolve/debug this > issue. > > Thanks, > Abhi > > > > -- > View this message in context: > http://activemq.2283324.n4.nabble.com/Consumer-not-able-to-consume-messages-from-queue-tp4689594p4689783.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. >
