[ 
https://issues.apache.org/jira/browse/QPIDJMS-430?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16740594#comment-16740594
 ] 

ASF GitHub Bot commented on QPIDJMS-430:
----------------------------------------

franz1981 commented on issue #26: QPIDJMS-430 Lock-Free FifoMessageQueue
URL: https://github.com/apache/qpid-jms/pull/26#issuecomment-453601630
 
 
   @tabish121 I haven't had the chance to measure the gains now that 
https://github.com/apache/activemq-artemis/pull/2467 has been merged and sadly 
I don't have such a fast box ATM, but I would try it and see if it can make any 
difference. 
   I can surely avoid the padding even if it will lower the performance, given 
that it would be anyway better then the orignal version, that was locked.
   @gemmellr @tabish121 My only worry is that the original version was using a 
resizable q of messages while this version is quite strict and never grows 
exceeding a configured capacity ie `prefetchSize`, I hope it is not a 
limitation, given that it would (yield)wait on offer side until one slot became 
available.
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


> Lock-Free FifoMessageQueue
> --------------------------
>
>                 Key: QPIDJMS-430
>                 URL: https://issues.apache.org/jira/browse/QPIDJMS-430
>             Project: Qpid JMS
>          Issue Type: Improvement
>            Reporter: Francesco Nigro
>            Priority: Minor
>
> A lock-free single producer multi consumer double ended queue would allow for 
> the common cases (enqueue/dequeue) to benefit from a better throughput and 
> lower latencies while allowing busy spin waits to benefits from some 
> additional boost



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

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

Reply via email to