[
https://issues.apache.org/jira/browse/MESOS-1442?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14032796#comment-14032796
]
Niklas Quarfot Nielsen commented on MESOS-1442:
-----------------------------------------------
Fixed in https://reviews.apache.org/r/22519/
> Adding grace_period_seconds to the health check protobuf
> --------------------------------------------------------
>
> Key: MESOS-1442
> URL: https://issues.apache.org/jira/browse/MESOS-1442
> Project: Mesos
> Issue Type: Technical task
> Components: master, slave
> Reporter: Niklas Quarfot Nielsen
> Assignee: Timothy Chen
>
> Using an initial delay to cope with start-up time of a task may be hard to
> get right in scenarios where this time varies a lot (for example, pulling in
> external artifacts). Overestimating the initial delay can be wasteful, as a
> task may be ready for work load much earlier. Instead, having a 'grace
> period' which would be interpreted as the slack time in which failed health
> checks won't count and cause a task to be killed, but would at first (or
> after a number of) healthy checks, report the task as being healthy.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
