On 21 May 2008, at 10:35, Artur Siekielski wrote:
Hi.
We use ActiveMQ 5.1.0 in the following way: we have four queues,
each queue
has about 5-12 consumers and producers, some of them are running on
different machines. Used transport is 'tcp'. Producers use non-
persistent
delivery mode, acknowledge mode is "auto". We do not create threads.
Messages are consumed synchronously using 'consumer.receive()'.
Exact number
of messages which have to be consumed/produced is known in advance.
Everything works ok for some time (like 0,5h), but after processing
eg. 500
000 messages all consumers of a specific queue hang. In admin
console there
are a few pending messages in that queue. How can we debug such
situation?
Turning DEBUG logging in broker didn't give us any useful information.
We have switched to latest ActiveMQ from 4.x branch and didn't
experience
hang situation yet (but more tests are needed).
Any help would be appreciated,
Artur
--
View this message in context:
http://www.nabble.com/ActiveMQ-5.x-hangs.-How-to-debug--tp17359508s2354p17359508.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Stack trace - and preferably a single junit test case - attached to a
jira would help identify the problem quicker
cheers,
Rob
http://open.iona.com/products/enterprise-activemq
http://rajdavies.blogspot.com/
