You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by "Dan Washusen (JIRA)" <ji...@apache.org> on 2010/02/24 11:20:28 UTC

[jira] Commented: (HBASE-2258) The WhileMatchFilter doesn't delegate the call to filterRow()

    [ https://issues.apache.org/jira/browse/HBASE-2258?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12837734#action_12837734 ] 

Dan Washusen commented on HBASE-2258:
-------------------------------------

Added bonus of fixing this issue is that the PE randomSeekScan test becomes dramatically faster.

Running the following before and after the fix yields:
{quote}
hbase/bin/hbase org.apache.hadoop.hbase.PerformanceEvaluation --rows=10 randomSeekScan 1
{quote}

Before: 198344ms at offset 0 for 10 rows
After: 238ms at offset 0 for 10 rows

> The WhileMatchFilter doesn't delegate the call to filterRow()
> -------------------------------------------------------------
>
>                 Key: HBASE-2258
>                 URL: https://issues.apache.org/jira/browse/HBASE-2258
>             Project: Hadoop HBase
>          Issue Type: Bug
>          Components: filters
>    Affects Versions: 0.20.3
>            Reporter: Dan Washusen
>             Fix For: 0.20.4
>
>
> While testing the recent MemStoreScanner slowness I noticed that each scan in the randomSeekScan test takes about 19 seconds to complete.  The scan in question provides a startRow and a WhileMatchFilter containing a PageFilter that asks for 120 rows.  I would have expected this scan to return in roughly the same amount of time as a scan that specifies a startRow and stopRow that spans a similar number of rows.
> As it turns out this is an issue with the WhileMatchFilter.  The WhileMatchFilter is not delegating the call the filterRow() down the the PageFilter.  As a result the PageFilter never increments the rowsAccepted counter.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.