Uploaded image for project: 'Giraph'
  1. Giraph
  2. GIRAPH-267

Jobs can get killed for not reporting status during INPUT SUPERSTEP

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: 1.0.0
    • Fix Version/s: 1.0.0
    • Component/s: graph
    • Labels:
      None
    • Environment:

      Facebook Hadoop

      Description

      Job with a skewed and long (>600secs in my case) INPUT_SUPERSTEP fails for some tasks not reporting their status. From BspServiceWorker#setup(), I could tell while some workers were still loading inputSplits, others finished theirs early and hanged on PredicateLock#waitForever(), and got killed after the timeout.

        Attachments

        1. GIRAPH-267.patch
          13 kB
          Jaeho Shin
        2. GIRAPH-267.patch
          13 kB
          Jaeho Shin
        3. 0001-Made-PredicateLock-report-progress-and-removed-Conte.patch
          11 kB
          Jaeho Shin

          Issue Links

            Activity

              People

              • Assignee:
                netj Jaeho Shin
                Reporter:
                netj Jaeho Shin
              • Votes:
                0 Vote for this issue
                Watchers:
                5 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: