Description of problem: Cluster-node-tuning-operator is crashing/restarting and is going though leader elections during the kube-apiserver rollout which takes around ~60 seconds with shutdown-delay-duration and gracefulTerminationDuration is now set to 0 and 15 seconds ( https://github.com/openshift/cluster-kube-apiserver-operator/pull/1168 and https://github.com/openshift/library-go/pull/1104 ). cluster-node-tuning-operator leader election timeout should be set to > 60 seconds to handle the downtime gracefully in SNO. Recommended lease duration values to be considered for reference as noted in https://github.com/openshift/enhancements/pull/832/files#diff-2e28754e69aa417e5b6d89e99e42f05bfb6330800fa823753383db1d170fbc2fR183: LeaseDuration=137s, RenewDealine=107s, RetryPeriod=26s. These are the configurable values in k8s.io/client-go based leases and controller-runtime exposes them. This gives us 1. clock skew tolerance == 30s 2. kube-apiserver downtime tolerance == 78s 3. worst non-graceful lease reacquisition == 163s 4. worst graceful lease reacquisition == 26s We can see that leader lease failures in the log: http://dell-r510-01.perf.lab.eng.rdu2.redhat.com/chaos/sno/cluster-node-tuning-operator/cluster-node-tuning-operator.log. The leader election can also be disabled given that there's no HA in SNO. Version-Release number of selected component (if applicable): 4.9.0-0.nightly-2021-07-26-031621 How reproducible: Always Steps to Reproduce: 1. Install a SNO cluster using the latest nightly payload. 2. Trigger kube-apiserver rollout/outage that lasts at least 60 seconds - $oc patch kubeapiserver/cluster --type merge -p '{"spec":{"forceRedeploymentReason":"ITERATIONX"}}' where X can be 1,2...n 3. Observe the state of cluster-node-tuning-operator. Actual results: cluster-node-tuning-operator is crashing/restarting and going through leader elections. Expected results: cluster-node-tuning-operator should handle the API rollout/outage gracefully. Additional info: Logs including must-gather: http://dell-r510-01.perf.lab.eng.rdu2.redhat.com/chaos/sno/cluster-node-tuning-operator/
Verified Steps: Cluster Version: 4.9.0-0.nightly-2021-08-02-145924 Test Steps: Startup SNO Cluster and delete api-server one by one and delete all api-server one time time oc delete pod apiserver-8568699655-kdlwj -n openshift-apiserver;time oc delete pod apiserver-8568699655-p2dhz -n openshift-apiserver;time oc delete pod apiserver-8568699655-vsl6z -n openshift-apiserver pod "apiserver-8568699655-kdlwj" deleted real 1m17.189s user 0m0.155s sys 0m0.063s pod "apiserver-8568699655-p2dhz" deleted real 1m21.086s user 0m0.107s sys 0m0.038s pod "apiserver-8568699655-vsl6z" deleted real 1m17.096s user 0m0.120s sys 0m0.024s Test Result: No POD restarted during api-server rollout oc get pods -n openshift-cluster-node-tuning-operator NAME READY STATUS RESTARTS AGE cluster-node-tuning-operator-b8d46fb9d-v8gzl 1/1 Running 0 62m tuned-9pwkq 1/1 Running 0 60m tuned-fmkjw 1/1 Running 0 60m tuned-gmpmn 1/1 Running 0 60m
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 (Moderate: OpenShift Container Platform 4.9.0 bug fix and security update), 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/RHSA-2021:3759