[jira] [Commented] (MESOS-3599) COMMAND health checks with marathon running in slave context broken

["Timothy Chen (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MESOS-3599?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15020885#comment-15020885 ] 

Timothy Chen commented on MESOS-3599:
-------------------------------------

Sorry joining this thread and related issues pretty late, didn't see there was all these discussions going on :)

I think my first question is that we didn't used to have any health check with Docker Containerizer, and when we added health checks it has always been running inside of the docker executor (thanks to Haosdent!). So I'm not really sure why moving back to 0.22.1 make it works, since your health check command is pretty much ignored with Docker Containerizer.

Now the ability to run outside of the container for a health check I think makes sense, I'm favoring more the first option Haosdent proposed since the ability to run inside of the container is also very important for dependency and lots of other reasons. We can ignore this option with http/tcp checks.

> COMMAND health checks with marathon running in slave context broken
> -------------------------------------------------------------------
>
>                 Key: MESOS-3599
>                 URL: https://issues.apache.org/jira/browse/MESOS-3599
>             Project: Mesos
>          Issue Type: Bug
>    Affects Versions: 0.23.0
>            Reporter: Erhan Kesken
>            Assignee: haosdent
>            Priority: Critical
>
> When deploying Mesos 0.23rc4 with latest Marathon 0.10.0 RC3 command health check stop working. Rolling back to Mesos 0.22.1 fixes the problem.
> Containerizer is Docker.
> All packages are from official Mesosphere Ubuntu 14.04 sources.
> The issue must be analyzed further.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)