[
https://issues.apache.org/jira/browse/MESOS-6676?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15770153#comment-15770153
]
Alexander Rukletsov commented on MESOS-6676:
--------------------------------------------
I've backported this to 1.1.1. [~vinodkone] I believe you still might want to
backport it to 1.0.x.
> Always re-link with scheduler during re-registration.
> -----------------------------------------------------
>
> Key: MESOS-6676
> URL: https://issues.apache.org/jira/browse/MESOS-6676
> Project: Mesos
> Issue Type: Bug
> Components: master
> Reporter: Neil Conway
> Assignee: Neil Conway
> Labels: mesosphere
> Fix For: 1.1.1, 1.2.0
>
>
> Scenario:
> # Framework registers with master using a non-zero {{failover_timeout}} and
> is assigned a FrameworkID.
> # The master sees an {{ExitedEvent}} for the master->scheduler link. This
> could happen due to some transient network error, e.g., 1-way partition. The
> master sends a {{FrameworkErrorMessage}} to the framework. The master marks
> the framework as disconnected, but keeps the {{Framework*}} for it around in
> {{frameworks.registered}}.
> # The framework doesn't receive the {{FrameworkErrorMessage}} because it is
> dropped by the network.
> # The scheduler might receive an {{ExitedEvent}} for the scheduler -> master
> link, but it ignores this anyway (see MESOS-887).
> # The scheduler sees a new-master-detected event and re-registers with the
> master. It doesn _not_ set the {{force}} flag. This means we follow [this
> code
> path|https://github.com/apache/mesos/blob/a6bab9015cd63121081495b8291635f386b95a92/src/master/master.cpp#L2771]
> in the master, which does _not_ relink with the scheduler.
> The result is that scheduler re-registration succeds, but the master ->
> scheduler link is never re-established.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
