Uploaded image for project: 'Hadoop YARN'
  1. Hadoop YARN
  2. YARN-7527

Over-allocate node resource in async-scheduling mode of CapacityScheduler

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Major
    • Resolution: Fixed
    • 3.0.0-alpha4, 2.9.1
    • 3.1.0, 2.10.0, 3.0.3
    • capacityscheduler
    • None
    • Reviewed

    Description

      Currently in async-scheduling mode of CapacityScheduler, node resource may be over-allocated since node resource check is ignored.
      FiCaSchedulerApp#commonCheckContainerAllocation will check whether this node have enough available resource for this proposal and return check result (ture/false), but this result is ignored in CapacityScheduler#accept as below.

      commonCheckContainerAllocation(allocation, schedulerContainer);
      

      If FiCaSchedulerApp#commonCheckContainerAllocation returns false, CapacityScheduler#accept should also return false as below:

      if (!commonCheckContainerAllocation(allocation, schedulerContainer)) {
        return false;
      }
      

      Attachments

        1. YARN-7527.001.patch
          6 kB
          Tao Yang
        2. YARN-7527-branch-2.001.patch
          6 kB
          Tao Yang
        3. YARN-7527-branch-2.002.patch
          6 kB
          Tao Yang

        Activity

          People

            Tao Yang Tao Yang
            Tao Yang Tao Yang
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: