Version-Release number of selected component (if applicable): 4.6.0-0.nightly-2020-08-23-185640 How reproducible: Always Steps to Reproduce: 1. Patch one node with egressCIDR oc patch hostsubnet compute-0 --type=merge -p '{"egressCIDRs": ["136.144.52.192/26"]}' 2. Create a ns and a pod in it, patch egressIP oc patch netnamespace test1 --type=merge -p '{"egressIPs": ["136.144.52.212"]}' netnamespace.network.openshift.io/test1 patched 3. Shutdown the egressIP node compute-0 Actual Result: oc get pods -n openshift-sdn -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES ovs-2k6zz 1/1 Running 0 149m 136.144.52.203 compute-1 <none> <none> ovs-4q565 1/1 Running 0 158m 136.144.52.196 control-plane-2 <none> <none> ovs-7w2nz 1/1 Running 0 158m 136.144.52.198 control-plane-0 <none> <none> ovs-g9wtr 1/1 Running 0 149m 136.144.52.204 compute-0 <none> <none> ovs-ps9dt 1/1 Running 0 158m 136.144.52.199 control-plane-1 <none> <none> sdn-9c4cz 1/1 Running 0 158m 136.144.52.198 control-plane-0 <none> <none> sdn-controller-24znf 0/1 CrashLoopBackOff 8 158m 136.144.52.198 control-plane-0 <none> <none> sdn-controller-fdnpj 0/1 CrashLoopBackOff 5 158m 136.144.52.196 control-plane-2 <none> <none> sdn-controller-w8w9k 1/1 Running 6 158m 136.144.52.199 control-plane-1 <none> <none> sdn-g6nd5 1/1 Running 0 158m 136.144.52.199 control-plane-1 <none> <none> sdn-k5skc 1/1 Running 0 149m 136.144.52.204 compute-0 <none> <none> sdn-metrics-cjxwk 1/1 Running 0 147m 136.144.52.203 compute-1 <none> <none> sdn-metrics-xbrvp 1/1 Running 0 147m 136.144.52.204 compute-0 <none> <none> sdn-s2tp5 1/1 Running 0 149m 136.144.52.203 compute-1 <none> <none> sdn-sxvjv 1/1 Running 0 158m 136.144.52.196 control-plane-2 <none> <none> oc logs sdn-controller-w8w9k -n openshift-sdn I0824 06:09:47.994332 1 leaderelection.go:242] attempting to acquire leader lease openshift-sdn/openshift-network-controller... E0824 06:09:48.018610 1 event.go:316] Could not construct reference to: '&v1.ConfigMap{TypeMeta:v1.TypeMeta{Kind:"", APIVersion:""}, ObjectMeta:v1.ObjectMeta{Name:"openshift-network-controller", GenerateName:"", Namespace:"openshift-sdn", SelfLink:"/api/v1/namespaces/openshift-sdn/configmaps/openshift-network-controller", UID:"e964ce50-2c57-41c6-8a61-a19ce7363673", ResourceVersion:"130002", Generation:0, CreationTimestamp:v1.Time{Time:time.Time{wall:0x0, ext:63733837208, loc:(*time.Location)(0x27fb000)}}, DeletionTimestamp:(*v1.Time)(nil), DeletionGracePeriodSeconds:(*int64)(nil), Labels:map[string]string(nil), Annotations:map[string]string{"control-plane.alpha.kubernetes.io/leader":"{\"holderIdentity\":\"control-plane-1\",\"leaseDurationSeconds\":60,\"acquireTime\":\"2020-08-24T06:08:11Z\",\"renewTime\":\"2020-08-24T06:09:47Z\",\"leaderTransitions\":1}"}, OwnerReferences:[]v1.OwnerReference(nil), Finalizers:[]string(nil), ClusterName:"", ManagedFields:[]v1.ManagedFieldsEntry{v1.ManagedFieldsEntry{Manager:"openshift-sdn-controller", Operation:"Update", APIVersion:"v1", Time:(*v1.Time)(0xc000408660), FieldsType:"FieldsV1", FieldsV1:(*v1.FieldsV1)(0xc0004086a0)}}}, Immutable:(*bool)(nil), Data:map[string]string(nil), BinaryData:map[string][]uint8(nil)}' due to: 'no kind is registered for the type v1.ConfigMap in scheme "pkg/runtime/scheme.go:101"'. Will not report event: 'Normal' 'LeaderElection' 'control-plane-1 became leader' I0824 06:09:48.018721 1 leaderelection.go:252] successfully acquired lease openshift-sdn/openshift-network-controller I0824 06:09:48.022210 1 master.go:52] Initializing SDN master I0824 06:09:48.034824 1 network_controller.go:61] Started OpenShift Network Controller I0824 06:09:48.034909 1 reflector.go:175] Starting reflector *v1.NetNamespace (10m0s) from k8s.io/client-go.2/tools/cache/reflector.go:125 I0824 06:09:48.034937 1 reflector.go:175] Starting reflector *v1.Namespace (10m0s) from k8s.io/client-go.2/tools/cache/reflector.go:125 I0824 06:09:48.034956 1 reflector.go:175] Starting reflector *v1.HostSubnet (10m0s) from k8s.io/client-go.2/tools/cache/reflector.go:125 I0824 06:09:48.035100 1 reflector.go:175] Starting reflector *v1.Node (10m0s) from k8s.io/client-go.2/tools/cache/reflector.go:125 panic: runtime error: invalid memory address or nil pointer dereference [signal SIGSEGV: segmentation violation code=0x1 addr=0x20 pc=0x14f217e] goroutine 127 [running]: github.com/openshift/sdn/pkg/network/master.(*egressIPManager).check(0xc000898360, 0x48331c00, 0x27fb000, 0xc0000b87e0, 0x0) github.com/openshift/sdn/pkg/network/master/egressip.go:180 +0x10e github.com/openshift/sdn/pkg/network/master.(*egressIPManager).poll(0xc000898360, 0xc0000b8f60) github.com/openshift/sdn/pkg/network/master/egressip.go:155 +0x78 created by github.com/openshift/sdn/pkg/network/master.(*egressIPManager).maybeDoUpdateEgressCIDRs github.com/openshift/sdn/pkg/network/master/egressip.go:127 +0x3b8 It seems caused by PR https://github.com/openshift/sdn/pull/171/files, from above logs, line 180,155 is the new codes in above PR. Expected Result: SDN pods are in running status and egressIP works well
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.6 GA Images), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4196