[
https://issues.apache.org/jira/browse/HBASE-7835?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13631447#comment-13631447
]
Jeffrey Zhong commented on HBASE-7835:
--------------------------------------
[~jxiang] Thanks for the reviewing! We still have the optimization. You can
check file HLogSplitter.java
{code}sequenceIdChecker.getLastSequenceId(serverNameStr, region);{code}.
Basically we record lastFlushedSequenceId in ZK and use them to skip flushed
edits. We store the information per RS to handle the scenario when a RS fails
multiple times in a row.
> Implementation of the log splitting without creating intermediate files
> ------------------------------------------------------------------------
>
> Key: HBASE-7835
> URL: https://issues.apache.org/jira/browse/HBASE-7835
> Project: HBase
> Issue Type: Sub-task
> Components: regionserver
> Reporter: Jeffrey Zhong
> Assignee: Jeffrey Zhong
> Fix For: 0.95.1
>
> Attachments: hbase-7006_12.patch, hbase-7006_13.patch,
> hbase-7006_1.patch, hbase-7006_2.patch, hbase-7006_3.patch,
> hbase-7006_3.patch, hbase-7006_6.patch, hbase-7006_8.patch
>
>
> The sub task is to check in a separate patch for major implementation for the
> proposal.
> More checkins will include a new "replay" command and more metrics support.
> Thanks,
> -Jeffrey
--
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
