[ 
https://issues.apache.org/jira/browse/CASSANDRA-3938?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13213104#comment-13213104
 ] 

Jonathan Ellis commented on CASSANDRA-3938:
-------------------------------------------

Is this really a performance pain point?  Doing a couple box/unboxes per 
sstable on CL replay doesn't seem like a hot spot to me.  Unless performance is 
a problem I think the meaning of the code is more clear as it is now.
                
> Optimize autoboxing in ReplayPosition.comparator
> ------------------------------------------------
>
>                 Key: CASSANDRA-3938
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-3938
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core
>    Affects Versions: 1.0.7, 1.1.0
>            Reporter: Jason Brown
>            Priority: Trivial
>              Labels: commitlog, replay
>             Fix For: 1.1.1
>
>         Attachments: 0001.ReplayPosition-autobox.patch
>
>
> ReplayPosition.comparator does some unnecessary boxing (and auto-boxing) from 
> primitive int and long to Integer and Long, just to perform a basic numeric 
> comparison. We can just inline the comparison and avoid the conversion to 
> Integer/Long instances. The ReplayPosition.equals() simple looks at the 
> primitive values, so this change would just make the code more parallel.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to