Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-36704

PTP operator showing high cpu utilization with steady workload on OCP 4.15

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done-Errata
    • Icon: Undefined Undefined
    • None
    • 4.15.z
    • Networking / ptp
    • Important
    • No
    • False
    • Hide

      None

      Show
      None
    • Release Note Not Required
    • In Progress
    • 6/20: dev will triage once a lab is available

      This is a clone of issue OCPBUGS-35804. The following is the description of the original issue:
      —
      Description of problem:

      
On SNO spoke deployed with RAN DU profile and PTP with BC profile,

PTP is showing higher cpu utilization during steady workload state than 4.14 and 4.16
(linuxptp-daemon pod)

The issue seems to have started appearing around the OCP  4.15.14 time frame

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

      OCP 4.15.18
PTP Operator version: 4.15.0-202406132106

      How reproducible:

      always since 4.15.14

      Steps to Reproduce:

      1. Deploy SNO spoke with Telco RAN DU profile
2. Run cpu_util 1h regression test. 
3. Observe per pod cpu utilization during steady workload phase

      Actual results:

      Almost a full CPU is consumed, 4.14 and 4.16 results are much lower.

      Expected results:

      PTP cpu utilization is comparable to other releases.

      Additional info:

      Splunk results comparison The chart of interest is "Avg cpu usage of Infra Pods Components by ocp build greater then 0.005 CPUs" in the left column

      Addtional debug info will be attached when node is available with 4.16.z deployed.

            aputtur@redhat.com Aneesh Puttur
            openshift-crt-jira-prow OpenShift Prow Bot
            Bonnie Block Bonnie Block
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

              Created:
              Updated:
              Resolved: