I can check the 5.3 snapshot, but I don't see the relationship between the
two leaks.

In my scenario, the consumers and sessions aren't meant to be reclaimed;
they're meant to be long-lived. The problem is that they seem to be hanging
onto a couple of Statistics objects for each message they handle.



Gary Tully wrote:
> 
> this leak sounds a lot like
> https://issues.apache.org/activemq/browse/AMQ-2178 that is fixed on trunk
> and available in the current 5.3-snapshot.
> 
> 2009/5/16 emersonf <emerson.farru...@gmail.com>
> 
>>
>> Hi,
>>
>> I've been running some load tests on a Spring Integration system backed
>> by
>> ActiveMQ 5.2.0. I noticed that one of my application components was
>> leaking
>> memory, and heap analysis showed this memory to be
>> org.apache.activemq.management statistics related. In particular,
>> instances
>> of CountStatisticImpl and TimeStatisticImpl are being created at an
>> alarming
>> rate, and they're not being garbage collected.
>>
>> The pertinent behavior of Spring Integration
>> (DefaultMessageListenerContainer, for those who care) is that it creates
>> a
>> fixed number of sessions on a single connection, and then a fixed number
>> of
>> message consumers on those sessions. These sessions and consumers are
>> kept
>> for the lifetime of the application. A message producer is created for
>> each
>> outgoing message.
>>
>> I have already disabled statistics on the broker
>> (enableStatistics="false"),
>> but the problem persists. Is there any way to disable statistics on the
>> broker's clients?
>>
>> Thanks,
>> Emerson
>> --
>> View this message in context:
>> http://www.nabble.com/Possible-memory-leak-in-statistics-gathering-tp23572764p23572764.html
>> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>>
>>
> 
> 
> -- 
> http://blog.garytully.com
> 
> Open Source SOA
> http://FUSESource.com
> 
> 

-- 
View this message in context: 
http://www.nabble.com/Possible-memory-leak-in-statistics-gathering-tp23572764p23606221.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Reply via email to