[
https://issues.apache.org/jira/browse/HBASE-3573?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-3573:
-------------------------
Attachment: 3573.txt
Here's a patch that removes the shutdown from HMsg and instead has the master
just wait till only catalog regions online. I thought this would be enough but
it isn't; current shutdown has RSs close regions w/o updating ZK so master is
never 'informed' about the closed regions.
Next up is a patch that does what we currently do via heartbeat; i.e. master
just waits till <=2 servers but instead of telling remote RS shutdown via
heartbeat, will add new graceful_shutdown RCP... we need it for rolling restart
anyways.
> Move shutdown messaging OFF hearbeat; prereq for fix of hbase-1502
> ------------------------------------------------------------------
>
> Key: HBASE-3573
> URL: https://issues.apache.org/jira/browse/HBASE-3573
> Project: HBase
> Issue Type: Task
> Reporter: stack
> Assignee: stack
> Fix For: 0.92.0
>
> Attachments: 3573.txt
>
>
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
