[
https://issues.apache.org/jira/browse/AURORA-821?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14164261#comment-14164261
]
Bill Farner commented on AURORA-821:
------------------------------------
The description doesn't give us much to work with. Can you include logs that
show the RPCs made to the scheduler, perhaps followed by an RPC result that
shows the modes of these hosts? The scheduler itself is storing the state, so
a failover should not cause state changes.
> Aurora_admin does not persist DRAINED state across scheduler fail-over
> ----------------------------------------------------------------------
>
> Key: AURORA-821
> URL: https://issues.apache.org/jira/browse/AURORA-821
> Project: Aurora
> Issue Type: Story
> Reporter: Tobias Weingartner
>
> State before fail-over:
> {noformat}
> 159 DRAINED
> 1 DRAINING
> 340 SCHEDULED
> {noformat}
> State after fail-over:
> {noformat}
> 500 SCHEDULED
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
