hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Matteo Bertozzi (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HBASE-8418) Caching KeyValue.getRowLength() to improve scan performance
Date Wed, 24 Apr 2013 15:13:16 GMT

     [ https://issues.apache.org/jira/browse/HBASE-8418?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Matteo Bertozzi updated HBASE-8418:
-----------------------------------

    Attachment: HBASE-8418-0.94-v0.patch

attached a 94 version of the patch that caches the getRowLength() and uses that instead of
toShort().

I still have to see if it really improves the performance, I keep getting different results
but my guess is because I'm running on virtual machines...

the profiler at least shows less calls to Bytes.toShort() that was the top function after
StoreScanner.next()
                
> Caching KeyValue.getRowLength() to improve scan performance
> -----------------------------------------------------------
>
>                 Key: HBASE-8418
>                 URL: https://issues.apache.org/jira/browse/HBASE-8418
>             Project: HBase
>          Issue Type: Improvement
>          Components: regionserver
>            Reporter: Matteo Bertozzi
>            Assignee: Matteo Bertozzi
>            Priority: Minor
>         Attachments: HBASE-8418-0.94-v0.patch
>
>
> Looking at the profile seems that Bytes.toShort() is called a lot.
> the calls are in ScanQueryMatcher.match() and KeyValue.KeyComparator.compare()
> getKeyLength() is already cached, but getRowLength() is not.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message