Thanks for the response ceposta. Yes, I get the reason for the flow control kicking in. My primary question though is, is there a way to make the broker A (producer) from resuming the publishing automatically once the consumer (broker B) has eased up instead of having to restart the broker A manually? As I mentioned broker A seems to continue storing the messages on disk even after broker B has recovered.
Once the brokerB was restarted, I see the transport connectivity being re-established between broker A and B in the broker A log, but no resumption of the publishing: 2013-02-05 02:26:49,751 [idge] Task-3000] INFO HashIndex- Resize hash bins queue-data_queue://Performance from 1024 to 2048 2013-02-05 15:14:02,642 [idge] Task-5685] INFO HashIndex- Resize hash bins queue-data_queue://Performance from 2048 to 4096 2013-02-06 16:48:58,929 [dge] Task-11054] INFO HashIndex- Resize hash bins queue-data_queue://Performance from 4096 to 8192 2013-02-08 19:47:48,090 [dge] Task-21934] INFO HashIndex- Resize hash bins queue-data_queue://Performance from 8192 to 16384 2013-02-11 09:33:17,875 [92.168.1.1:6117] WARN FailoverTransport - Transport (/brokerBHostName:6117) failed to tcp://brokerBHostName:6117 , attempting to automatically reconnect due to: java.io.EOFException 2013-02-11 09:33:18,831 [92.168.1.1:6116] WARN FailoverTransport - Transport (/brokerBHostName:6116) failed to tcp://brokerBHostName:6116 , attempting to automatically reconnect due to: java.io.EOFException *2013-02-11 09:34:28,898 [iveMQ Task-4586] INFO FailoverTransport - Successfully reconnected to tcp://brokerBHostName:6117 2013-02-11 09:34:29,827 [iveMQ Task-4585] INFO FailoverTransport - Successfully reconnected to tcp://brokerBHostName:6116* -- View this message in context: http://activemq.2283324.n4.nabble.com/Producer-broker-never-resumes-publishing-after-ResourceAllocationException-tp4663276p4663282.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
