Hello, when testing Apache Active MQ 5.3 and 5.4 we run in several issues regarding the cleanup of logfiles with KahaDB as persistence layer. We are now concerned about this issue appearing on Apache Active MQ 5.5 under specific circumstance. Therefore we want to have a workaround if this problem reappeared to avoid message loss in production.
As from my current understanding transaction and queue messages are stored within the log files and which log file stores the message for which queue is not transparent, our current solution is as follows: - Read for each queue all messages and store them in a file - Purge all queues - Stop Active MQ Broker - Delete all files under data/ directory - Start Active MQ Broker - Write for each queue all messages from file to the queue Are there any other workarounds, tools or something like that which I can use in this scenario? Where can I find more information about the Kaha DB? Can I externally trigger/force the cleanup in some ways? Thanks for any help -- View this message in context: http://activemq.2283324.n4.nabble.com/kaha-db-store-structure-and-workarounds-in-production-tp4302787p4302787.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
