Hi, It appears we may be getting by this https://issues.apache.org/activemq/browse/AMQ-2009 or some variant of it i.e. messages are getting stuck in a queue. We are busy looking at our config and gradually making changes to try and resolve it but we can't replicate in dev so it's slow going. We're using ActiveMQ 5.2.0 in a network of brokers.
Once a number of messages are stuck on a queue, we're using a couple of workarounds to get them resuscitated: Bounce the brokers. This usually gets reinvigorates about 50% of them. The rest become immune. Copy the messages off the queue using a query browser, purge the original queue and resubmit the messages. This works but there's a real danger of losing a message that arrived in the interim. Keep doing the above until none of the messages are stuck. It would be great if there was a way, from the WebConsole, to shake the messages awake, btw. Anyone else faced with a similar problem? How have you solved it? Our latest configuration change attempt is to reduce the prefetch down to 1 for all of our consumers. regards, Jamie.
