Uploaded image for project: 'Hadoop Map/Reduce'
  1. Hadoop Map/Reduce
  2. MAPREDUCE-7327

Job.waitForCompletion function can sleep most for 596 hours when jobclient.completion.poll.interval is misconfigured , causing the job to hang

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Open
    • Major
    • Resolution: Unresolved
    • 3.3.0
    • None
    • client
    • None

    Description

      The loop terminates depending on a configurable value and there is little sanity checking on this value. When jobclient.completion.poll.interval is misconfigured to INT_MAX, it can cause the loop to sleep at most for 596 hours. The thread would get stuck and never return to the user even if the job completes. We suggest adding a cap value or a warning message.

      public boolean waitForCompletion(boolean verbose
                                         ) throws IOException, InterruptedException,
                                                  ClassNotFoundException {
      ... 
         while (!isComplete()) {
              try {
                  Thread.sleep(completionPollIntervalMillis);
              } catch (InterruptedException ie) {
          }
      ...
      }
      

      Attachments

        Activity

          People

            Unassigned Unassigned
            tshan Tina Shan
            Votes:
            1 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated: