[
https://issues.apache.org/jira/browse/MESOS-1475?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14036911#comment-14036911
]
Benjamin Mahler commented on MESOS-1475:
----------------------------------------
The first part (SIGUSR1 signal handler) is committed:
{noformat}
commit 671b4c4b852dab10645af5e94fcdacde48c54adf
Author: Alexandra Sava <[email protected]>
Date: Wed Jun 18 20:14:39 2014 -0700
Added a SIGUSR1 handler to cleanly shut down the slave.
Review: https://reviews.apache.org/r/21379/
commit aaef6bcf9be799dda82b207c6f6cb1cbaf02f0fc
Author: Alexandra Sava <[email protected]>
Date: Wed Jun 18 18:07:06 2014 -0700
Added a SIGUSR1 shutdown test.
Review: https://reviews.apache.org/r/22737
{noformat}
> Provide a way to fully shut down a slave (kill all tasks underneath).
> ---------------------------------------------------------------------
>
> Key: MESOS-1475
> URL: https://issues.apache.org/jira/browse/MESOS-1475
> Project: Mesos
> Issue Type: Improvement
> Components: master, slave
> Reporter: Benjamin Mahler
> Assignee: Alexandra Sava
> Labels: gsoc2014
>
> Per [~tweingartner] on MESOS-544. For maintenance operations, there needs to
> be a way to fully shutdown a slave in a manner that kills all the underlying
> tasks.
> This can be accomplished using a SIGUSR1 handler:
> {noformat}
> # drain current host
> pkill -USR1 mesos-slave
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.2#6252)
