[
https://issues.apache.org/jira/browse/HBASE-7390?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13536570#comment-13536570
]
stack commented on HBASE-7390:
------------------------------
bq. But on master failover, when we iterate on the znodes, we ignore the one
int the SPLIT state, that may be the reason.
Yes. In case the master went away on us, so the new master gets the SPLIT
callback...and its handler does the SPLIT cleanup.
Comment doesn't enough:
{code}
* Transitions an existing node for the specified region which is
* currently in the SPLITTING state to be in the SPLIT state. Converts the
* ephemeral SPLITTING znode to an ephemeral SPLIT node. Master cleans up
* SPLIT znode when it reads it (or if we crash, zk will clean it up).
{code}
> Add extra test cases for assignement on the region server and fix the related
> issues
> ------------------------------------------------------------------------------------
>
> Key: HBASE-7390
> URL: https://issues.apache.org/jira/browse/HBASE-7390
> Project: HBase
> Issue Type: Bug
> Components: Region Assignment, regionserver
> Affects Versions: 0.96.0
> Reporter: nkeywal
> Assignee: nkeywal
> Fix For: 0.96.0
>
> Attachments: 7390.v1.patch, 7390.v2.patch, 7390.v3.patch,
> 7390.v4.patch, assignment_zk_states.jpg
>
>
> We don't have a lot of tests on the region server itself.
> Here are some.
> Some of them are failing, feedback welcome.
> See as well the attached state diagram for the ZK nodes on assignment.
--
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
