Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-18662

cnf-tests: [test_id: 55012] RPS configuration applied on some physical devices

    XMLWordPrintable

Details

    • +
    • No
    • CNF Compute Sprint 246
    • 1
    • False
    • Hide

      None

      Show
      None
    • Hide
       Previously, a race condition between udev events and the creation queues associated with physical device led to some of the queues being configured with the wrong Receive Packet Steering (RPS) mask while they should be reset to zero. This resulted in the RPS mask being configured on the queues of the physical devices, meaning they were using RPS instead of Receive Side Scaling (RSS) which could impact the performance. With this fix, the event was changed to be triggered per queue creation instead of on device creation. This guarantees that no queue will be missing. The queues of all physical devices are now set up with the correct RPS mask which is empty. (link:https://issues.redhat.com/browse/OCPBUGS-18662[*OCPBUGS-18662*])
      Show
       Previously, a race condition between udev events and the creation queues associated with physical device led to some of the queues being configured with the wrong Receive Packet Steering (RPS) mask while they should be reset to zero. This resulted in the RPS mask being configured on the queues of the physical devices, meaning they were using RPS instead of Receive Side Scaling (RSS) which could impact the performance. With this fix, the event was changed to be triggered per queue creation instead of on device creation. This guarantees that no queue will be missing. The queues of all physical devices are now set up with the correct RPS mask which is empty. (link: https://issues.redhat.com/browse/OCPBUGS-18662 [* OCPBUGS-18662 *])
    • Bug Fix
    • In Progress
    • 9/26: telco priority for 4.14 pending triage

    Description

      Description of problem:
      RPS configuration test failed with the following error:

      [FAILED] Failure recorded during attempt 1:
      a host device rps mask is different from the reserved CPUs; have "0" want ""
      Expected
          <bool>: false
      to be true
      In [It] at: /tmp/cnf-ZdGbI/cnf-features-deploy/vendor/github.com/onsi/gomega/internal/assertion.go:62 @ 09/06/23 03:47:44.144
      < Exit [It] [test_id:55012] Should have the correct RPS configuration - /tmp/cnf-ZdGbI/cnf-features-deploy/vendor/github.com/openshift/cluster-node-tuning-operator/test/e2e/performanceprofile/functests/1_performance/performance.go:337 @ 09/06/23 03:47:44.144 (39.949s)
      

      Full report: 

      https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/origin-ci-test/logs/periodic-ci-openshift-release-master-nightly-4.15-e2e-telco5g-cnftests/1699249554244767744/artifacts/e2e-telco5g-cnftests/telco5g-cnf-tests/artifacts/test_results.html 

      How reproducible:

      Very often

      Steps to Reproduce:
      1. Reproduce automatically by the cnf-tests nightly job

      Actual results:
      Some of the virtual devices are not configured with the correct RPS mask

      Expected results:
      All virtual network devices are expected to have the correct RPS mask

      Attachments

        Issue Links

          Activity

            People

              titzhak Talor Itzhak
              titzhak Talor Itzhak
              Mallapadi Niranjan Mallapadi Niranjan
              Votes:
              0 Vote for this issue
              Watchers:
              11 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: