activemq-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ARTEMIS-1156) Long Autoboxing occurring on Hot Path
Date Wed, 10 May 2017 07:51:04 GMT

    [ https://issues.apache.org/jira/browse/ARTEMIS-1156?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16004238#comment-16004238
] 

ASF GitHub Bot commented on ARTEMIS-1156:
-----------------------------------------

Github user michaelandrepearce commented on the issue:

    https://github.com/apache/activemq-artemis/pull/1263
  
    @franz1981 
    On 1) compute isn't being used within this class, the methods used are ::get ::put ::contains
::containsKey, indeed this is why i state this is targeted, instead of your original blanket
replace all, each case should be evaluated and implemented on merit.
    
    On 2) agreed, in warm up stages we would see this, but once warm and message rate through
the broker, it would stabilise its size, this is why we don't see hashmap.entry being created
heavily in the profiler once warm.
    
    On 3) good spot. lets address that.


> Long Autoboxing occurring on Hot Path
> -------------------------------------
>
>                 Key: ARTEMIS-1156
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-1156
>             Project: ActiveMQ Artemis
>          Issue Type: Improvement
>          Components: Broker
>    Affects Versions: 2.0.0
>            Reporter: Michael Andre Pearce (IG)
>         Attachments: Screen Shot 2017-05-10 at 02.05.47.png
>
>
> JournalImpl is on the hot path, and whilst a lot of effort seems to have gone into ensuring
the use of primitives, unfortunately due to java collections used it is forcing long id's
to be auto boxed, this causes un-needed memory pressures. 



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message