[
https://issues.apache.org/jira/browse/HBASE-14687?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14972156#comment-14972156
]
Elliott Clark commented on HBASE-14687:
---------------------------------------
Yeah and even the javadoc for BufferdMutator makes it sound like HTable where
there should be one per thread. That ends up being really really bad as there's
one AsyncProcess per buffered mutator. This can lead to a single regionserver
being pounded.
However if you don't use more than one buffered mutator then the locking is
extreme and contention eats up your cpu.
> Un-synchronize BufferedMutator
> ------------------------------
>
> Key: HBASE-14687
> URL: https://issues.apache.org/jira/browse/HBASE-14687
> Project: HBase
> Issue Type: Bug
> Components: Client, Performance
> Reporter: Elliott Clark
> Assignee: Elliott Clark
>
> It should totally be possible to make BufferedMutatorImpl not use much ( if
> any ) locking.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
