[
https://issues.apache.org/activemq/browse/SM-1631?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=46426#action_46426
]
Guillaume Nodet commented on SM-1631:
-------------------------------------
Well, I guess one workaround would be to make sure the threads are released.
This can be done by changing the values in [install]/conf/servicemix.properties
to the following values:
{code}
servicemix.corePoolSize = 0
servicemix.maximumPoolSize = 32
servicemix.queueSize = 0
{code}
After 60 seconds, all unused threads will die so the references to the
SAXParsers will go away and the GC can reclaim the used memory.
> Memory leak while using eip component (wiretab, static recipient)
> -----------------------------------------------------------------
>
> Key: SM-1631
> URL: https://issues.apache.org/activemq/browse/SM-1631
> Project: ServiceMix
> Issue Type: Bug
> Affects Versions: servicemix-eip-2008.01
> Reporter: Steffen Rumpf
> Priority: Blocker
> Attachments: FileToThreeFiles.png, LongRun.png,
> servicemix-bean-test.zip, servicemix-memory-leak.png,
> servicemix-memory-leak.zip
>
>
> I am using a eip-component (tested both static recipient and wiretab) and
> received a memory leak while processing messages. There is a simple testcase
> for this, when receiving a file and send it via eip to more then one file
> sender there is a remaining amount of heap which is not cleared by gc (even
> when I trigger the gc by hand).
> Related threads:
> http://www.nabble.com/Best-practices%2C-working-with-beans-and-huge-messages.-tt19914439ef12049.html
> http://www.nabble.com/Memory-issue-in-servicemix-tt19727632ef12049.html
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
