[
https://issues.apache.org/jira/browse/VYSPER-332?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13489983#comment-13489983
]
Emmanuel Lecharny commented on VYSPER-332:
------------------------------------------
It *could* happen if the client does not read the messages fast enough: MINA
stacks all the messages in a list and wait until the client's socket is ready
for OP_WRITE. In some cases, I have seen this list growing and growing up to a
point we get an OOM.
Bottom line: the Handler should check that the list is not too big before
writting some new messages using the session.getWriteRequestQueue().size()
method.
Now, this is a problem I faced in another MINA based application, not sure it
applies to Vysper.
> Memory Leak in BoshBackedSessionContext
> ---------------------------------------
>
> Key: VYSPER-332
> URL: https://issues.apache.org/jira/browse/VYSPER-332
> Project: VYSPER
> Issue Type: Bug
> Components: BOSH
> Affects Versions: 0.7
> Environment: - Ubuntu 12.04.1 LTS (GNU/Linux 3.2.0-31-virtual x86_64)
> - OpenJDK 64-Bit Server VM (build 23.2-b09, mixed mode)
> - Jetty 6.1.24
> Reporter: Patrick Twohig
>
> I'm seeing the heap size grow disproportionately large when running the
> BoshServlet in my web application. Specifically, I am seeing the consume
> 300+ megabytes of heap causing an OOME. The heap dump indicates that roughly
> 85% of my memory is consumed but the single instanced of . I can provide a
> heap dump if needs be.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
