Uploaded image for project: 'Aurora'
  1. Aurora
  2. AURORA-788

GC executor doesn't need to exit after period of inactivity

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Minor
    • Resolution: Fixed
    • None
    • 0.6.0
    • None
    • None
    • Aurora Q4 Sprint 1

    Description

      The GC Executor is currently configured to exit after a hardcoded limit of activity since it processed a launchTask request. This causes needless task churn in the cluster, especially given that the scheduler's default launch interval will cause the executor to need to be relaunched on average every 15 minutes.

      Attachments

        Issue Links

          relates to

          Bug - A problem which impairs or prevents the functions of the product. MESOS-243 driver stop() should block until outstanding requests have been persisted

          • Major - Major loss of function.
          • Resolved

          Activity

            People

              kevints Kevin Sweeney
              kevints Kevin Sweeney
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: