Artemis 2.6.3 running out of memory with slow consumer and fast producer

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Artemis 2.6.3 running out of memory with slow consumer and fast producer

kkaczkow
Hi,

Currently we are performing Artemis crash testing before we switch
completely to it from ActiveMQ. We've applied following scenario:
- Artemis broker 2.6.3 with heap size of 2Gb, global-max-size of 400Mb, and
global address-setting to apply DROP policy whenever memory exceeds 150Mb
(max-size-bytes).
- Two slow consumers consuming every 10 sec
- Two fast producers sending messages to 2000 topics every 100 ms each

The broker is running out of memory in a couple of hours of testing. Would
you have any idea if I misconfigured anything? I tried also applying PAGE
strategy with similar effect. I would expect the broker to drop messages
once memory size exceeds 400Mb.

I've also noticed similar issue:
https://issues.apache.org/jira/browse/ARTEMIS-2164, do you think it might be
related?

Thanks for your help!
Konrad



--
Sent from: http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html
Reply | Threaded
Open this post in threaded view
|

Re: Artemis 2.6.3 running out of memory with slow consumer and fast producer

kkaczkow
Actually, I realized this problem can be simplified to having just two fast
producers, even without any consumers at all. I am sending messages using
ActiveMQ 5.15.2 JMS client.
The issue occurs also on artemis 2.7.0-SNAPSHOT version.



--
Sent from: http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html