You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Lars Hofhansl (JIRA)" <ji...@apache.org> on 2013/04/20 08:41:17 UTC

[jira] [Comment Edited] (HBASE-8362) Possible MultiGet optimization

    [ https://issues.apache.org/jira/browse/HBASE-8362?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13637138#comment-13637138 ] 

Lars Hofhansl edited comment on HBASE-8362 at 4/20/13 6:40 AM:
---------------------------------------------------------------

Why can't StoreFileScanner.requestSeek make use of a ROW bloom filter? We'd check the ROW BF before the ROWCOL BF and if is does not pass we can immediately hint to the last KV on the row. That way if a row is not found in an HFile at all, we can skip that file, rather looking for all columns.
(It's probably not that simple, or we would do that already)

Edit: NM, I guess we would skip that HFile anyway with the key range check, so a ROW BF should never match.
                
      was (Author: lhofhansl):
    Why can't StoreFileScanner.requestSeek make use of a ROW bloom filter? We'd check the ROW BF before thew ROWCOL BF and it is does not pass the ROW BF we can immediately hint to the last KV on the row. That way if a row is not found in an HFile at all, we can skip that file, rather looking for all columns.
(It's probably not that simple, or we would do that already)

                  
> Possible MultiGet optimization
> ------------------------------
>
>                 Key: HBASE-8362
>                 URL: https://issues.apache.org/jira/browse/HBASE-8362
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Lars Hofhansl
>
> Currently MultiGets are executed on a RegionServer in a single thread in a loop that handles each Get separately (opening a scanner, seeking, etc).
> It seems we could optimize this (per region at least) by opening a single scanner and issue a reseek for each Get that was requested.
> I have not tested this yet and no patch, but I would like to solicit feedback on this idea.

--
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