You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Nicolas Spiegelberg (JIRA)" <ji...@apache.org> on 2012/04/25 00:02:06 UTC

[jira] [Commented] (HBASE-5867) Improve Compaction Throttle Default

    [ https://issues.apache.org/jira/browse/HBASE-5867?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13261066#comment-13261066 ] 

Nicolas Spiegelberg commented on HBASE-5867:
--------------------------------------------

The most common type of compaction is compacting only flushed files.  Assuming that there is no compression (the default), then the common compaction size should be:
{code}
minFiles * flushSize
{code}
The current idea is to support this operation and supporting a compaction with 1 previously-compacted file.  Assuming no overlap, this size would be: (minFiles-1) * flushSize + minFiles * flushSize ==>
{code}
2 * minFiles * flushSize - ε
{code}
                
> Improve Compaction Throttle Default
> -----------------------------------
>
>                 Key: HBASE-5867
>                 URL: https://issues.apache.org/jira/browse/HBASE-5867
>             Project: HBase
>          Issue Type: Improvement
>            Reporter: Nicolas Spiegelberg
>            Assignee: Nicolas Spiegelberg
>            Priority: Minor
>
> We recently had a production issue where our compactions fell behind because our compaction throttle was improperly tuned and accidentally upgraded all compactions to the large pool.  The default from HBASE-3877 makes 1 bad assumption: the default number of flushed files in a compaction.  Currently the algorithm is:
> throttleSize ~= flushSize * 2
> This assumes that the basic compaction utilizes 3 files and that all 3 files are compressed.  In this case, "hbase.hstore.compaction.min" == 6 && the values were not very compressible.  Both conditions should be taken into consideration.  As a default, it is less damaging for the large thread to be slightly higher than it needs to be versus having everything accidentally promoted.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira