Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-44644

[release-4.17] PPC: false negative reporting while comparing the topologies of affected compute nodes

XMLWordPrintable

    • Moderate
    • None
    • CNF Compute Sprint 262, CNF Compute Sprint 263, CNF Compute Sprint 264
    • 3
    • False
    • Hide

      None

      Show
      None
    • Hide
      * Previously, the validation of Non-Uniform Memory Access (NUMA) node topologies expected equal core indices. This expectation caused the Performance Profile Creator (PPC) to report inaccurate information for compute nodes. With this release, the validation check no longer expects equal core indices so that the PPC now reports accurate information for compute nodes. (link:https://issues.redhat.com/browse/OCPBUGS-44644[*OCPBUGS-44644*])
      Show
      * Previously, the validation of Non-Uniform Memory Access (NUMA) node topologies expected equal core indices. This expectation caused the Performance Profile Creator (PPC) to report inaccurate information for compute nodes. With this release, the validation check no longer expects equal core indices so that the PPC now reports accurate information for compute nodes. (link: https://issues.redhat.com/browse/OCPBUGS-44644 [* OCPBUGS-44644 *])
    • Bug Fix
    • In Progress
    • Nov 8: malformed code spotted; currently working on a reproducer internally. A workaround to unblock the business can be building the performanceprofile manually.

      This is a clone of issue OCPBUGS-44372. The following is the description of the original issue:
      —
      Description of problem:

         This bug is filed a result of https://access.redhat.com/support/cases/#/case/03977446
ALthough both nodes topologies are equavilent, PPC reported a false negative:

  Error: targeted nodes differ: nodes host1.development.lab and host2.development.lab have different topology: the CPU corres differ: processor core #20 (2 threads), logical processors [2 66] vs processor core #20 (2 threads), logical processors [2 66]

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

          4.14

      How reproducible:

          always

      Steps to Reproduce:

          1.TBD
    2.
    3.

      Actual results:

          Error: targeted nodes differ: nodes host1.development.lab and host2.development.lab have different topology: the CPU corres differ: processor core #20 (2 threads), logical processors [2 66] vs processor core #20 (2 threads), logical processors [2 66]

      Expected results:

          topologies matches, the PPC should work fine

      Additional info:

              rhn-support-shajmakh Shereen Haj
              openshift-crt-jira-prow OpenShift Prow Bot
              Mallapadi Niranjan Mallapadi Niranjan
              Votes:
              0 Vote for this issue
              Watchers:
              5 Start watching this issue

                Created:
                Updated: