[
https://issues.apache.org/jira/browse/HBASE-3504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12990475#comment-12990475
]
dhruba borthakur commented on HBASE-3504:
-----------------------------------------
The HDFS write call is already synchronized within the HDFS client. So the
HBase code does not need the updateLock for writing into a HDFS file handle.
The reason HBase needs the updateLock is to ensure that the current writer is
not closed (by the log roller) when it is being used to sync/write by another
thread. The other operations like the invocation to obtainSeqNum() and
updating lastSeqWritten does not need to be protected by the updateLock, isn't
it?
> HLog performance improvement
> ----------------------------
>
> Key: HBASE-3504
> URL: https://issues.apache.org/jira/browse/HBASE-3504
> Project: HBase
> Issue Type: Improvement
> Components: regionserver
> Reporter: dhruba borthakur
> Assignee: dhruba borthakur
>
> The HLog.updateLock protects the rolling of logs with concurrent writes to
> the HDFS log file. This is a scalability bottleneck for a workload that
> comprises mostly of counter-increments.
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
