hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Billy Pearson (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-917) filesystem intensive operations such as compaction should be load aware
Date Fri, 10 Oct 2008 09:26:44 GMT

    [ https://issues.apache.org/jira/browse/HBASE-917?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12638513#action_12638513
] 

Billy Pearson commented on HBASE-917:
-------------------------------------

I see my compaction stop now during a long insert jobs that is loading the server vary heavy.
I thank it related to HBASE-616 but I do not get any messages from in the logs as we see from
HBASE-616
Once insert load slows the next memcache flush starts a compaction.

> filesystem intensive operations such as compaction should be load aware
> -----------------------------------------------------------------------
>
>                 Key: HBASE-917
>                 URL: https://issues.apache.org/jira/browse/HBASE-917
>             Project: Hadoop HBase
>          Issue Type: Improvement
>          Components: regionserver
>            Reporter: Andrew Purtell
>
> If the underlying filesystem is already severely stressed, running intensive operations
such as compaction is asking for trouble. Ideally, such actions should be deferred until load
is observed to lessen. 

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


Mime
View raw message