Are the messages that are redelivered only ones that were consumed in the
seconds before the original master came down? Could this simply be a
situation where message acknowledgements are being buffered in memory prior
to being written to disk and the process died before the disk write
occurred? Or
I have a use case where in there exist 2 broker connected to same kahadb.
In case of failover i can see that one of the broker is activated
sucessfully. However i notice that message in queue are re-processed i.e.
there are duplicate message
I am using active mq 5.15.0
appreciate if any one c
