[
https://issues.apache.org/jira/browse/COTTON-1?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14653082#comment-14653082
]
Yan Xu commented on COTTON-1:
-----------------------------
https://reviews.apache.org/r/37063/
https://reviews.apache.org/r/37064/
> Scheduler should recover only once
> ----------------------------------
>
> Key: COTTON-1
> URL: https://issues.apache.org/jira/browse/COTTON-1
> Project: COTTON
> Issue Type: Bug
> Reporter: Yan Xu
> Assignee: Yan Xu
> Original Estimate: 1m
> Remaining Estimate: 1m
>
> Currently the scheduler recovers its state every time {{registered()}} is
> called, however until until https://issues.apache.org/jira/browse/MESOS-2910
> is fixed / released {{registered()}} could be called when the scheduler
> reconnects to a failed over Mesos master, in which case we don't want the
> scheduler to recover again (and this messes up with scheduler metrics).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
