-
Bug
-
Resolution: Done
-
Undefined
-
None
-
None
-
Incidents & Support
-
False
-
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Upgrade to https://amd64.ocp.releases.ci.openshift.org/releasestream/4.21.0-0.nightly/release/4.21.0-0.nightly-2025-10-02-215712 from https://amd64.ocp.releases.ci.openshift.org/releasestream/4.21.0-0.nightly/release/4.21.0-0.nightly-2025-10-02-002727
: [Monitor:legacy-test-framework-invariants-pathological][sig-arch] events should not repeat pathologically for ns/openshift-kube-apiserver-operator { 1 events happened too frequently event happened 25 times, something is wrong: namespace/openshift-kube-apiserver-operator deployment/kube-apiserver-operator hmsg/6de5680bef - reason/OperatorVersionChanged clusteroperator/kube-apiserver version "operator" changed from "4.21.0-0.nightly-2025-10-02-002727" to "4.21.0-0.nightly-2025-10-02-215712" (23:02:22Z) result=reject }
We suspect https://github.com/openshift/origin/pull/30272 which modified the pathological events test and first landed in 4.21.0-0.nightly-2025-10-02-215712
We are suspecting a change in some kind of allowlist that makes these surface as failures, because these high event counts can be observed in older jobs, but they dot not trip the failure. Example from Sep 09: https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/logs/periodic-ci-openshift-release-master-ci-4.21-e2e-aws-upgrade-ovn-single-node/1965161661731966976/artifacts/e2e-aws-upgrade-ovn-single-node/gather-must-gather/artifacts/event-filter.html