Uploaded image for project: 'Marathon'
  1. Marathon
  2. MARATHON-7515

Document the TASK_UNREACHABLE/unreachableStrategy behavior of Marathon

    Details

      Description

      We should document on DC/OS level how the Mesos agent flags

      • --maxSlavePingTimeouts
      • --slavePingTimeout
      • --agent_reregister_timeout

      affect the TASK_UNREACHABLE behavior and timings, as well as the handling of the "unreachableStrategy" settings in the Marathon REST API (see https://github.com/mesosphere/marathon/blob/master/docs/docs/rest-api/public/api/v2/schema/AppDefinition.json#L816).

      Baseline is that the behavior is driven by the Mesos agent settings, and not what is defined in the Marathon app definitions. Currently, this can lead to confusion.

      IMO, this should be covered on https://mesosphere.github.io/marathon/docs/configure-task-handling.html 

        Attachments

          Issue Links

            Activity

              People

              • Assignee:
                suzanne Suzanne Scala (Inactive)
                Reporter:
                tmueller Tobias Mueller
                Team:
                Documentation Team
                Watchers:
                Suzanne Scala (Inactive), Tobias Mueller
              • Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: