My ActiveMQ broker instance is configured for a recovery policy to allow messages (up to 6 minutes old) to be resent even after acknowledgement. If the client app is restarted, it will receive redelivered messages (if they were brokered within the past 6 minutes). This works great, except for one thing...
It appears that *if* a new message is brokered during the down-time of the client app's restart, only this new message is delivered (after the client app is back in business), and nothing older (but < 6 mins) is redelivered. If there were no new messages brokered during the client's restart, then the client app gets all the messages of the past 6 minutes, i.e., it works fine! The behavior I desire is that once my client is back up, all messages within the last 6 minutes are delivered *followed by* the new one that was created during the down-time. I looked through all the configuration options shown on the ActiveMQ web site, but no configuration setting stood out as being a potential solution to this problem. Why would a new message brokered during the down-time prevent ActiveMQ from honoring the recovery policy when it *does* honor the recovery policy when there were no new brokered messages?? Mark -- View this message in context: http://activemq.2283324.n4.nabble.com/ActiveMQ-not-always-honoring-Recovery-Policy-tp4655052.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
