hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "James Taylor (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-7805) Scan of same region processed serially if table has region observer coprocessor
Date Mon, 11 Feb 2013 02:17:12 GMT

    [ https://issues.apache.org/jira/browse/HBASE-7805?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13575597#comment-13575597

James Taylor commented on HBASE-7805:

The coprocessor time will be much faster - they're running the scan completely on the server
side and only returning a single row back to the client (mimicing what Phoenix does for an
ungrouped aggregate query). Still not using the enhancements that Lars added, so the number
will get better after that. 

The 0.94.2 numbers are fine, but my coprocessor wasn't being used for the 0.94.5 run. The
jar path needs to be specified as null or else the coprocessor is silently ignored. The updated
number are below (nice improvement on the non coprocessor case for 0.94.5):

10,000,000 rows
v 0.94.2 (@1395367)
With coprocessors
Serial time: 5352, Parallel time: 3548 (66%)
Without coprocessors
Serial time: 13442, Parallel time: 7456 (55%)

v 0.94.5 (latest)
With coprocessors
Serial time: 5089, Parallel time: 3604 (71%)
Without coprocessors
Serial time: 9711, Parallel time: 6226 (64%)
> Scan of same region processed serially if table has region observer coprocessor
> -------------------------------------------------------------------------------
>                 Key: HBASE-7805
>                 URL: https://issues.apache.org/jira/browse/HBASE-7805
>             Project: HBase
>          Issue Type: Bug
>          Components: Coprocessors, Scanners
>    Affects Versions: 0.94.4
>            Reporter: James Taylor
>            Priority: Minor
>             Fix For: 0.96.0, 0.94.6
>         Attachments: 7805-0.94.patch, 7805-trunk.txt, 7805-trunk-v2.txt, ParallelScanTestCoProc.java,
> If multiple scans to different parts of the same region are executed in parallel, they
are processed serially if the table has a region observer coprocessor.

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

View raw message