Hi, When reviewing ActiveMQ code, I see a lot of special treatment for the prefetch=0 case. What exactly is the conceptual difference between 1 and 0 prefetch?
We are currently using prefetch=0 via failover, but are in the process of migrating away from failover to multiple connections with Spring DefaultMessageListenerContainer, which won't stop on container shutdown, as the message is stopped before the MessageDispatch from the broker is returned (i.e. there is the special case that the MessageConsumer waits indefinitely for a MessagePull from the server on prefetch=0). So what exactly do I have to expect as difference in regards to message ordering, load balancing, message groups when using prefetch=1 instead of prefetch=0? Thanks for any explanation! Best regards, Martin
