[
https://issues.apache.org/jira/browse/QPID-3366?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Justin Ross closed QPID-3366.
-----------------------------
Resolution: Won't Fix
Fix Version/s: JIRA Cleanup
This issue has not been updated in more than a year. We've reviewed
it and decided to close it, as part of an effort to keep our issue
database clean and current. If this issue is still affecting you,
please reopen it.
See https://issues.apache.org/jira/browse/QPID-4585 for more
information.
> qpid get stuck under heavy load
> -------------------------------
>
> Key: QPID-3366
> URL: https://issues.apache.org/jira/browse/QPID-3366
> Project: Qpid
> Issue Type: Bug
> Components: C++ Client
> Affects Versions: 0.10
> Environment: Linux
> Reporter: Francesco Emmi
> Fix For: JIRA Cleanup
>
>
> Hi all,
> I' ve been experiencing a very strange behaviour on qpid C++ client that
> seems to arise just only under heavy load.
> Basically in the queue named "frames" inside client::TCPConnector::writebuf
> somehow a frame arrives which is grater than 65535 bytes. This cause
> client::TCPConnector::encode not to be able to transform the frame in a
> AsinchIO buffer and therefore the frame is not pop-ed from frames queue while
> on the other side e 0 sized buffer is inserted into AynchIO buffers queue.
> Starting from this moment on, the application using qpid client get stuck no
> longer sending messages and using 100% CPU. This is becouse the big frame in
> front of frames queue cannot be elimiated and acts like a cork in the queue.
> I' ve no idea so far how to handle this, just because I cannot understand
> who creates this big frame and how to prevent this to happen.
> I'll continue to investigate and try to give you as more details as possible.
> Thanks
> Francesco
--
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
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
