I’ve had this problem before . . . A supplier is working on a system Cloverleaf is sending messages to and this has resulted in a large number of messages building up in the outbound queue. I’ve checked with ‘top’ and the engine is taking in excess of 95% cpu. I have to freeze the cluster otherwise cpu usage over 95% for more than five minutes triggers a failover.
Why does the engine use so much cpu when messages build up in the queue and can I do anything to mitigate this?
Dave