[
https://issues.apache.org/jira/browse/AXIS-2668?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499805
]
Prasanna Sundarrajan commented on AXIS-2668:
--------------------------------------------
Hi Bjorn,
Yes, we tried with Axis 1.4 also, there also the behaviour is consistent.
As per our application requirement the volume of data involved is large(around
3 to 4 MB) in few transaction and we found that all these objects are being
held by org/apache/axis/MessageContext objects. These MessageContext objects
are inturn being held by java/lang/ref/Finalizer objects.
So we are not sure whether the finalizers are executed and hence the
MessageContext objects will not be removed from the heap.
(MessageContext.finalize() method)
If the user performing the same transaction(transaction involves 4MB of data)
for around 30 minutes then we are getting outofmemory issue. Is there any other
alternative to this problem, can we force the finalizer by running
"System.runFinalization" externally? Any idea or any other pointers? Kindly let
me know.
Thanks and Regards
Prasanna
> Axis 1.3 - Garbage collection issue
> -----------------------------------
>
> Key: AXIS-2668
> URL: https://issues.apache.org/jira/browse/AXIS-2668
> Project: Axis
> Issue Type: Bug
> Affects Versions: 1.3
> Environment: OS - AIX Application Server Oracle Applicatio Server,
> AXIS Version 1.3
> Reporter: Prasanna Sundarrajan
> Priority: Critical
>
> Hi All,
> This is Prasanna. Currently I am working for an application which is a
> service oriented application.
> The services are designed using Java - Axis 1.3 and the services are hosted
> by the Oracle Application
> Server running under AIX (IBM JDK version 1.5) and the services are consumed
> by .net application.
> During our performance load testing we found a weird behavior in our
> application. We would like to share the same with you. It would be great if
> any of you kindly look in to the below issue and give any solution to resolve
> the same.
> We had encountered Out-of memory exception in
> the java application server when a single user performs the same transaction
> repeatedly for more than 30 minutes. We tested this scenario using
> performance
> load scripts. We have analyzed the heap dump and found that the objects
> created in the application server are not garbage collected and they still
> have references and all the references are created under MessageContext
> objects.
> As part of this exercise we profiled the java code and none of the code has
> memory leaks.
> We would like to know how this to be handled do we need to close or dispose
> anything explicitly. Looking forward your valuable response. Any help or
> pointer is highly appreciated.
> Thanks in advance
> Prasanna.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
