activemq-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hiram Chirino (JIRA)" <>
Subject [jira] [Commented] (APLO-160) Apollo becoming unresponsive when stressed with 48k connections.
Date Mon, 30 Apr 2012 14:55:49 GMT


Hiram Chirino commented on APLO-160:

Hi Lionel,

Several changes have recently been added which should help keep memory levels down in large
connection scenarios:

r1332222 Producers were holding on to large 64k buffers needlessly.
r1331585 keep it simple. copy all the subscribe headers.
r1331577 Deep copy the subscription id to avoid holding onto a large buffer.
r1331565 Use smarter buffer pooling in the StompCodec so that idle connections don't hog memory
r1331491 Don't initialize the read/write buffers until we know what size to make them.
r1331470 Use the new transport aware interface of hawtdispatch to more reliably figure out
the read/write buffer sizes for the stomp codec.

Please try out a new nightly SNAPSHOT, you should see a much improved memory profile.
> Apollo becoming unresponsive when stressed with 48k connections.
> ----------------------------------------------------------------
>                 Key: APLO-160
>                 URL:
>             Project: ActiveMQ Apollo
>          Issue Type: Bug
>         Environment: apollo-1.1-20120209.032648-24
>            Reporter: Lionel Cons
>            Assignee: Hiram Chirino
>             Fix For: 1.3
>         Attachments: apollo.dump
> While running a stress test against apollo-1.1-20120209.032648-24 (many concurrent TCP
connections), the broker became unresponsive.
> It logged several times: java.lang.OutOfMemoryError: GC overhead limit exceeded
> It also logged other warnings, probably related:
> 2012-02-14 14:14:49,273 | WARN  | handle failed | | Apollo Task
> 2012-02-14 14:18:39,073 | WARN  | Problem scavenging sessions | org.eclipse.jetty.server.session
| HashSessionScavenger-0
> It could not be stopped either, I had to kill -9 it.
> What can be done to avoid these problems?
> FWIW, java has been started with -server -Xmx8192m -Xms4096m

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


View raw message