You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "ramkrishna.s.vasudevan (JIRA)" <ji...@apache.org> on 2013/08/05 21:38:48 UTC

[jira] [Commented] (HBASE-9118) Optimization in HFile V3 when no tags are present in a file

    [ https://issues.apache.org/jira/browse/HBASE-9118?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13729857#comment-13729857 ] 

ramkrishna.s.vasudevan commented on HBASE-9118:
-----------------------------------------------

We could use HBASE-9127(something similar) and refactor how mvcc works along with the tags optimization.
                
> Optimization in HFile V3 when no tags are present in a file
> -----------------------------------------------------------
>
>                 Key: HBASE-9118
>                 URL: https://issues.apache.org/jira/browse/HBASE-9118
>             Project: HBase
>          Issue Type: Sub-task
>    Affects Versions: 0.98.0, 0.95.2
>            Reporter: Anoop Sam John
>            Assignee: Anoop Sam John
>
> Now with V3 we will write tags length (short) along with every KV after value bytes. When no tags at all present it will be wasting 2 bytes for every KV. We can avoid this. During flush let the tags length be written. In fileInfo we can add info like max tags length. During compaction if all the files undergoing the compaction having a 0 max tags length, we can avoid writing 2 bytes (0) along with every KV..
> Note : Similar optimization available with mvcc

--
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