[
https://issues.apache.org/jira/browse/HBASE-609?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-609:
------------------------
Status: Patch Available (was: Open)
Please review.
I tried this patch on our test cluster. I put the master 3 minutes behind the
regionservers. Doing this, I was able to replicate the issue reported by the
streamy fellas: ISE region offline. Applying the patch and rerunning upload,
error went away.
> Master doesn't see regionserver edits because of clock skew
> -----------------------------------------------------------
>
> Key: HBASE-609
> URL: https://issues.apache.org/jira/browse/HBASE-609
> Project: Hadoop HBase
> Issue Type: Bug
> Reporter: stack
> Priority: Blocker
> Fix For: 0.2.0, 0.1.2
>
> Attachments: skew.patch
>
>
> The streamy folks had a cluster where regionserver was 2 minutes in advance
> of the master. On split, regionserver would update .META. with split info
> but scanners opened on the master wouldn't see the edits because they were
> being opened using current time.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
