Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-18490

CI fails on "[sig-arch][Late] operators should not create watch channels very often"

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done-Errata
    • Icon: Undefined Undefined
    • 4.12.z
    • 4.12.z
    • Monitoring
    • None
    • -
    • Low
    • No
    • MON Sprint 243
    • 1
    • Rejected
    • False
    • Hide

      None

      Show
      None
    • N/A
    • Release Note Not Required

      This is a clone of the below bug for 4.12 since https://search.ci.openshift.org/?search=Operator+%5C%5C%22prometheus-operator%5C%5C%22+produces+more+watch+requests+than+expected&maxAge=168h&context=1&type=junit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job returns results for 4.12 as well.

      Description of problem

      4.11 CI is flaky because of test failures such as the following:

      {  fail [github.com/openshift/origin/test/extended/apiserver/api_requests.go:449]: Expected
          <[]string | len:1, cap:1>: [
              "Operator \"prometheus-operator\" produces more watch requests than expected: watchrequestcount=197, upperbound=180, ratio=1.0944444444444446",
          ]
      to be empty
      Ginkgo exit error 1: exit with code 1}
      

      This particular failure comes from https://prow.ci.openshift.org/view/gs/origin-ci-test/pr-logs/pull/openshift_cluster-ingress-operator/894/pull-ci-openshift-cluster-ingress-operator-release-4.11-e2e-aws-single-node/1642909565664104448. Search.ci has additional similar errors.

      Version-Release number of selected component (if applicable)

      I have seen these failures in 4.11 CI jobs.

      How reproducible

      Presently, search.ci shows the following stats for the past 7 days:

      Found in 0.01% of runs (0.05% of failures) across 139252 total runs and 7354 jobs (16.18% failed)
      

      Steps to Reproduce

      1. Post a PR and have bad luck.
      2. Check search.ci: https://search.ci.openshift.org/?search=Operator+%5C%5C%22prometheus-operator%5C%5C%22+produces+more+watch+requests+than+expected&maxAge=168h&context=1&type=junit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

      Actual results

      CI fails.

      Expected results

      CI passes, or fails on some other test failure, and the failures don't show up in search.ci.

      Additional info

      This issue is not blocking my PR; the affected job, e2e-aws-single-node for cluster-dns-operator's release-4.11 branch, is optional. Please prioritize this bug report as you see fit.

              jfajersk@redhat.com Jan Fajerski
              mmasters1@redhat.com Miciah Masters
              Junqi Zhao Junqi Zhao
              Pranshu Srivastava
              Votes:
              0 Vote for this issue
              Watchers:
              5 Start watching this issue

                Created:
                Updated:
                Resolved: