-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
4.17
cluster-manager logs a failure to update the POD annotation and succeeds on a retry:
I0805 11:10:27.447621 1 kube.go:315] Updating pod e2e-test-endpointslices-mirror-e2e-7zbdf/backend-1 I0805 11:10:27.497062 1 network_cluster_controller.go:360] Pod update failed for e2e-test-endpointslices-mirror-e2e-7zbdf/backend-1, will try again later: failed to update pod e2e-test-endpointslices-mirror-e2e-7zbdf/backend-1: admission webhook "pod.network-node-identity.openshift.io" denied the request: user: "system:serviceaccount:openshift-ovn-kubernetes:ovn-kubernetes-control-plane" is not allowed to set k8s.ovn.org/pod-networks on pod "backend-1": no ovn pod annotation for network default: "{\"e2e-test-endpointslices-mirror-e2e-7zbdf/gryffindor\":{\"ip_addresses\":[\"203.203.0.2/16\",\"2014:100:200::2/60\"],\"mac_address\":\"0a:58:cb:cb:00:02\",\"tunnel_id\":2,\"role\":\"secondary\"}}" I0805 11:10:27.500607 1 kube.go:315] Updating pod e2e-test-endpointslices-mirror-e2e-7zbdf/backend-1 I0805 11:10:27.560224 1 pod.go:62] [e2e-test-endpointslices-mirror-e2e-7zbdf/backend-1] pod update took 59.64693ms
This looks like a race we should investigate.