[ 
https://issues.apache.org/jira/browse/QPIDJMS-565?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17502883#comment-17502883
 ] 

Robbie Gemmell commented on QPIDJMS-565:
----------------------------------------

Note that the original referenced change is actually from 3 years ago but that 
it didn't affect the 4.1.x releases so far. However it is now being backported 
and so will be included into the coming _4.1.75_ release by 
[https://github.com/netty/netty/pull/12109] , which will introduce a change in 
behaviour even for the EventLoop (as opposed to just 'other threads') if it is 
not using a FastThreadLocalThread.

> QpidJMSThreadFactory should create FastThreadLocalThread as Netty event loop 
> threads
> ------------------------------------------------------------------------------------
>
>                 Key: QPIDJMS-565
>                 URL: https://issues.apache.org/jira/browse/QPIDJMS-565
>             Project: Qpid JMS
>          Issue Type: Bug
>            Reporter: Francesco Nigro
>            Priority: Major
>
> [https://github.com/netty/netty/pull/8991] is causing thread local Netty 
> pooled buffers to *not* be used for allocations within 
> non-FastThreadLocalThread threads.
> It should be better to use FastThreadLocalThread instead and save this 
> performance issue to happen.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to