Thanks Gastón. Do we have a shepherd for this? It's not critical but make it hard to distinguish task killed by failed health check from killed by scheduler.
wt., 13.12.2016 o 16:55 użytkownik Gastón Kleiman <gas...@mesosphere.io> napisał: > On Mon, Dec 12, 2016 at 12:47 PM, Tomek Janiszewski <jani...@gmail.com> > wrote: > > > It there any information that kill is the result of failed healthcheck? > > TaskHealthStatus should have some details on what was wrong. When default > > executor is killing task it should add a reason and details to > TaskStatus. > > What do you think? > > > > I agree with you, so I created > https://issues.apache.org/jira/browse/MESOS-6786 with some ideas. > > -Gastón >
