Memory leak in Spring BlazeDS Integration 1.0.3.RELEASE due to AsyncMessage

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

Memory leak in Spring BlazeDS Integration 1.0.3.RELEASE due to AsyncMessage

VascoAce
I'm getting an AsyncMessage memory leak. This leak fills the Tomcat JVM tenured space after a few hours of heavy load operation and it crashes with an out of memory (OOM) error. I'm using the Spring BlazeDS Integration 1.0.3.RELEASE., Spring 3.0.1, Tomcat 6.0.26, ActiveMQ 5.3.1, SUN JDK 1.6.0_20. I first supected a slow JVM garbage collector (GC) so I tried many conbination of JVM settings with no luck. I'm setting the AsyncMessage to null, set the message time to live to 5000 (5 seconds) , set the message timestamp, and the AsyncMessage keep having some weak references that is preventing the GC from collecting the objects. I also tried by not setting the message time to live. Have anybody found a workaround to this issue? I'm going to try with a nightly build of blazeds or the Spring BlazeDS Integration to see if this issue is solved. Any ideas please let me know. This bug is delaying the release of our Flex application. There is a Adobe Blazeds Jira issue already submitted about the same issue with Jprofiler snapshots:

http://bugs.adobe.com/jira/browse/BLZ-502

Thanks,

Alberto Acevedo

Reply | Threaded
Open this post in threaded view
|

Re: Memory leak in Spring BlazeDS Integration 1.0.3.RELEASE due to AsyncMessage

FlexCoders mailing list
I have the same problem now,

Memory leak due to a large amount of the memory heap space is allocated by the AsyncMessage Objects.

I'm using the BlazeDS 4, Spring 3.0.5, WebSphere 7, ActiveMQ 5.2, JDK 1.6.

Kindly advise.

Thanks,

Sherif
Reply | Threaded
Open this post in threaded view
|

Re: Memory leak in Spring BlazeDS Integration 1.0.3.RELEASE due to AsyncMessage

FlexCoders mailing list
In reply to this post by VascoAce
I have the same problem now,

Memory leak due to a large amount of the memory heap space is allocated by
the AsyncMessage Objects.

I'm using the BlazeDS 4, Spring 3.0.5, WebSphere 7, ActiveMQ 5.2, JDK 1.6.

Kindly advise.

Thanks,

Sherif
---In [hidden email], <vascoAce@...> wrote :

 I'm getting an AsyncMessage memory leak. This leak fills the Tomcat JVM tenured space after a few hours of heavy load operation and it crashes with an out of memory (OOM) error. I'm using the Spring BlazeDS Integration 1.0.3.RELEASE., Spring 3.0.1, Tomcat 6.0.26, ActiveMQ 5.3.1, SUN JDK 1.6.0_20. I first supected a slow JVM garbage collector (GC) so I tried many conbination of JVM settings with no luck. I'm setting the AsyncMessage to null, set the message time to live to 5000 (5 seconds) , set the message timestamp, and the AsyncMessage keep having some weak references that is preventing the GC from collecting the objects. I also tried by not setting the message time to live. Have anybody found a workaround to this issue? I'm going to try with a nightly build of blazeds or the Spring BlazeDS Integration to see if this issue is solved. Any ideas please let me know. This bug is delaying the release of our Flex application. There is a Adobe Blazeds Jira issue already submitted about the same issue with Jprofiler snapshots:
 
 http://bugs.adobe.com/jira/browse/BLZ-502 http://bugs.adobe.com/jira/browse/BLZ-502
 
 Thanks,
 
 Alberto Acevedo