Uploaded image for project: 'Hadoop YARN'
  1. Hadoop YARN
  2. YARN-3744

ResourceManager should avoid allocating AM to same node repeatedly in case of AM launch failures

    XMLWordPrintableJSON

    Details

    • Type: Improvement
    • Status: Resolved
    • Priority: Major
    • Resolution: Duplicate
    • Affects Version/s: None
    • Fix Version/s: None
    • Component/s: None
    • Labels:
      None

      Description

      We have seen that if AM launch fails on some node due to configuration or bad disk issue YARN-3591, quite often it gets reallocated on the same node, causing job failures if the AM attempt limit is reached.

      It would be preferable if the scheduler can try to allocate AM on different nodes for subsequent attempts

        Attachments

          Issue Links

            Activity

              People

              • Assignee:
                Unassigned
                Reporter:
                jaideepdhok Jaideep Dhok
              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: