[ 
https://issues.apache.org/jira/browse/MESOS-3829?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14990959#comment-14990959
 ] 

haosdent commented on MESOS-3829:
---------------------------------

DOCKER_STOP_TIMEOUT is used for 
http://docs.docker.com/engine/reference/commandline/stop/. According docker 
document, you could receive SIGTERM before SIGKILL.

> Error on gracefully shutdown task
> ---------------------------------
>
>                 Key: MESOS-3829
>                 URL: https://issues.apache.org/jira/browse/MESOS-3829
>             Project: Mesos
>          Issue Type: Bug
>    Affects Versions: 0.25.0
>         Environment: Marathon: 0.12.0-RC1
> Mesos: 0.25.0
> Docker 1.9.0
>            Reporter: Rafael Capucho
>
> Hello,
> I'm suffering from the same error reported here[1]. I have configured my 
> mesos-slave environment as [2] setting DOCKER_STOP_TIMEOUT and 
> EXECUTOR_SHUTDOWN_GRACE_PERIOD.
> When I see the sandbox stdout, I can see in the first line the declaration:
> --stop_timeout="30secs"
> properly configured, but when I click in "Destroy App" in Marathon the stdout 
> keep showing weird things[3] like repeatedly "Killing docker task Shutting 
> down".
> In my code I deal with SIGTERM and it isn't being reached.
> Thank you.
> [1] - 
> https://groups.google.com/forum/?hl=en#!topic/marathon-framework/Oy0dN0Lron0
> [2] - https://paste.ee/r/grRyS
> [3] - https://paste.ee/r/SghOr



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

Reply via email to