[
https://issues.apache.org/jira/browse/AMQ-7300?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16924379#comment-16924379
]
ASF subversion and git services commented on AMQ-7300:
------------------------------------------------------
Commit 65de327369f24de008f857211b97d3c69283a932 in activemq's branch
refs/heads/master from Christopher L. Shannon (cshannon)
[ https://gitbox.apache.org/repos/asf?p=activemq.git;h=65de327 ]
AMQ-7300 - Remove unnecessary call to sequence updater as read is already
volatile
> LongSequenceGenerator should use AtomicLongFieldUpdater for performance
> -----------------------------------------------------------------------
>
> Key: AMQ-7300
> URL: https://issues.apache.org/jira/browse/AMQ-7300
> Project: ActiveMQ
> Issue Type: Improvement
> Components: Broker
> Affects Versions: 5.15.10
> Reporter: Christopher L. Shannon
> Assignee: Christopher L. Shannon
> Priority: Minor
> Fix For: 5.16.0
>
>
> I was doing some performance testing and profiling a broker with a high
> message throughput and noticed a hot spot on the getNextSequenceId() method
> of the LongSequenceGenerator class. This class gets used a lot of generating
> sequences in the broker including messages and currently it's synchronization
> for the methods.
> Instead we can use an AtomicLongFieldUpdater and volatile long to remove the
> synchronization. Using an AtomicLongFieldUpdater instead of an AtomicLong
> will prevent having to garbage collect a lot of new objects.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
