[jira] [Updated] (AURORA-360) The scheduler should have authority on reacting to task health check failures

["Bill Farner (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/AURORA-360?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Bill Farner updated AURORA-360:
-------------------------------

    Component/s:     (was: Client)
                 Executor

> The scheduler should have authority on reacting to task health check failures
> -----------------------------------------------------------------------------
>
>                 Key: AURORA-360
>                 URL: https://issues.apache.org/jira/browse/AURORA-360
>             Project: Aurora
>          Issue Type: Story
>          Components: Executor, Scheduler
>            Reporter: Bill Farner
>
> When the executor is performing HTTP health checks, it makes the decision to kill tasks when health checks are failing.  This could have ill effects if an entire service starts failing health checks simultaneously.  To have a better chance at keeping a service up in the face of faulty or sensitive health checks, the scheduler should make the decision about reacting to health checks.



--
This message was sent by Atlassian JIRA
(v6.2#6252)