Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-52852

Distribution of 1000's of VMs over 9 nodes leaves 1 node empty

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Major Major
    • None
    • 4.16.z, 4.18.z
    • kube-scheduler
    • Quality / Stability / Reliability
    • False
    • Hide

      None

      Show
      None
    • 0.42
    • Important
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      Description of problem:

      Distribution of 1000's of VMs over 9 nodes leaves 1 node empty

      Version-Release number of selected component (if applicable):

      OCP 4.16.30
      CNV 4.16.6

      How reproducible:

      Unknown, can reproduce on current cluster.

      Steps to Reproduce:

      1. Allocate 9 nodes for VMs
      2. Create 1000's of running VMs (increase allowed pod-per-node count as necessary)
      3. Stop all VMs and subsequently restart them if result not immediately achieved

      Actual results:

      1 node receives no VMs

      Expected results:

      Relatively fair distribution between all 9 nodes

      Additional info:

      Confirmed the node can run the VMs by cordoning other 8 nodes.
      Behaves the same with or without defaultCPUModel set.
      Confirmed multiple matching CPU models between the nodes.
      On this cluster, the 1 node is always the same node.

       

              fdeutsch@redhat.com Fabian Deutsch
              rhn-support-sbennert Sarah Bennert
              None
              None
              Node Team Bot Account Node Team Bot Account
              None
              Votes:
              0 Vote for this issue
              Watchers:
              11 Start watching this issue

                Created:
                Updated: