Re: Message dispatch performance using ActiveMQ 5 on application using 4 quad-cores machines

Hi Rob, Many thanks for your reply. I've attached the stack-trace as an attachment. Cheers, David Rob Davies wrote: Hi David, Your application should be i/o bound with a pre-fetch of 0 (effectively pul every message from the broker) - so something else is going on Could you send a full sta

Re: Message dispatch performance using ActiveMQ 5 on application using 4 quad-cores machines

Hi David, Your application should be i/o bound with a pre-fetch of 0 (effectively pul every message from the broker) - so something else is going on Could you send a full stack trace ? thanks, Rob On Dec 20, 2007, at 6:55 AM, David Sitsky wrote: Hi, I have simplistically speaking, an ap

Message dispatch performance using ActiveMQ 5 on application using 4 quad-cores machines

Hi, I have simplistically speaking, an application using ActiveMQ 5 which has a queue called work-items. A master JVM process puts work items onto the queue (this process also runs the embedded JMS broker), and external worker JVM processes read these items using receive() with a timeout of