I have an application that sends requests through a (persistent) queue, with multiple machines as consumers that can process the requests. To send 1000 requests takes 40 seconds, which is very slow. Is there any way to speed this up?
Also... Some of the messages require just a little work and get processed in 1/100 seconds. Others may take a few minutes of processing time. Is there a way to get a large pre-fetch and do some kind of "put back" if the consumer determines it will be spending a few minutes on 1 request? My producers can tag the messages before submitting them with a "processing time estimate". Is there a way to evenly distribute the tasks to different consumers efficiently? (When I had a large pre-fetch, one consumer would possibly take 20 messages that required 2-3 minutes each; it would be better to only get one of those large messages, or lots of smaller messages) -- View this message in context: http://activemq.2283324.n4.nabble.com/Slow-and-fast-messages-tp3660295p3660295.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
