Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -480,6 +480,22 @@ func NewUniversalPathologicalEventMatchers(kubeConfig *rest.Config, finalInterva
jira: "https://issues.redhat.com/browse/OCPBUGS-69400",
})

// The kubelet calls getImageVolumes() on every SyncPod reconciliation, which emits a
// "Pulled" event for each image volume even when the pod is in steady state. Pods with
// many image volumes (e.g. capi-operator with 10 provider images) accumulate hundreds of
// these benign events over their lifetime. Remove when upstream fix lands:
// https://github.com/kubernetes/kubernetes/issues/138644
registry.AddPathologicalEventMatcherOrDie(&SimplePathologicalEventMatcher{
name: "ImageVolumeAlreadyPresent",
locatorKeyRegexes: map[monitorapi.LocatorKey]*regexp.Regexp{
monitorapi.LocatorNamespaceKey: regexp.MustCompile(`^openshift-cluster-api-operator$`),
monitorapi.LocatorPodKey: regexp.MustCompile(`^capi-operator-`),
},
messageReasonRegex: regexp.MustCompile(`^Pulled$`),
messageHumanRegex: regexp.MustCompile(`already present on machine`),
jira: "https://issues.redhat.com/browse/OCPBUGS-84659",
})

registry.AddPathologicalEventMatcherOrDie(AllowBackOffRestartingFailedContainer)

registry.AddPathologicalEventMatcherOrDie(AllowOVNReadiness)
Expand Down