[
https://issues.apache.org/jira/browse/HBASE-9457?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13762212#comment-13762212
]
stack commented on HBASE-9457:
------------------------------
[~jxiang] That would be good if we could get away w/ that. In the past we had
a 'historian' table into which we wrote all cluster events. We used to have
interesting issues where the historian was not available when user tables were
trying to report they were splitting or closing so we used to try and ensure
it was opened first and closed last. If we were to do the same thing today --
or have a metrics system table -- then we'd want it up always and up before
user tables came on line... but maybe that is too hard to ensure? And maybe
stepped assign is for post 0.96? i.e. we should undo the special casing of
namespace system table? [~toffer]?
> Master could fail start if region server with system table is down
> ------------------------------------------------------------------
>
> Key: HBASE-9457
> URL: https://issues.apache.org/jira/browse/HBASE-9457
> Project: HBase
> Issue Type: Bug
> Components: master, Region Assignment
> Reporter: Jimmy Xiang
> Assignee: Jimmy Xiang
> Priority: Critical
>
> In the region server holding the system table is killed while master is
> starting, master will hang there waiting for system table to be assigned
> which won't happen.
--
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
