Uploaded image for project: 'Apache Twill'
  1. Apache Twill
  2. TWILL-46

Have a way to specify / control the restart action upon container failure.

    Details

    • Type: Improvement
    • Status: Resolved
    • Priority: Major
    • Resolution: Duplicate
    • Affects Version/s: 0.1.0-incubating
    • Fix Version/s: 0.10.0
    • Component/s: api, core
    • Labels:
      None

      Description

      Currently when a container exit abnormally, AM always restart it. It would be better if the application can have finer control. E.g. restarts up to N times.

        Issue Links

          Activity

          Hide
          chtyim Terence Yim added a comment -

          Not a direct way. You could, however, hit the application master resource REST endpoint to discover what runnables are running in what host and container (Click the Application Master link on the right in the YARN Dashboard, and append "/resources" the end of the URL, or through the TwillController.getResourceReport() API).

          Show
          chtyim Terence Yim added a comment - Not a direct way. You could, however, hit the application master resource REST endpoint to discover what runnables are running in what host and container (Click the Application Master link on the right in the YARN Dashboard, and append "/resources" the end of the URL, or through the TwillController.getResourceReport() API).
          Hide
          safderraza Safder Raza added a comment -

          I was looking into finding out if any container failed. First pass, I don't think there is a way. Please correct me if I am wrong.

          Do we have any updates in this issue? Thoughts?

          Show
          safderraza Safder Raza added a comment - I was looking into finding out if any container failed. First pass, I don't think there is a way. Please correct me if I am wrong. Do we have any updates in this issue? Thoughts?

            People

            • Assignee:
              jwang47 Alvin Wang
              Reporter:
              chtyim Terence Yim
            • Votes:
              1 Vote for this issue
              Watchers:
              4 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved:

                Development