Hide Forgot
Description of problem: After upgrade from 4.3 to 4.4 node tuning operator cannot get resources "node" in API group. Version-Release number of selected component (if applicable): 4.4 How reproducible: 100% Steps to Reproduce: 1. Create OCP 4.3 cluster. 2. Update to version 4.4 oc patch clusterversion/version --patch '{"spec":{"upstream":"https://openshift-release.svc.ci.openshift.org/graph"}}' --type=merge oc adm upgrade --to-image=registry.svc.ci.openshift.org/ocp/release:4.4.0-0.nightly-2020-01-16-103546 --allow-explicit-upgrade --force 3. Check logs in tuned pods. Actual results: On each node the same: I0116 18:16:04.422258 7008 openshift-tuned.go:550] Pod (openshift-kube-storage-version-migrator/migrator-76ff7f7965-6nz6g) labels changed node wide: true I0116 18:16:05.357354 7008 openshift-tuned.go:408] Dumping labels to /var/lib/tuned/ocp-pod-labels.cfg I0116 18:16:05.358560 7008 openshift-tuned.go:441] Getting recommended profile... I0116 18:16:05.451877 7008 openshift-tuned.go:638] Active and recommended profile (openshift-node) match. Label changes will not trigger profile reload. E0116 18:16:29.348558 7008 openshift-tuned.go:881] Error getting node ip-10-0-48-132.us-east-2.compute.internal: nodes "ip-10-0-48-132.us-east-2.compute.internal" is forbidden: User "system:serviceaccount:openshift-cluster-node-tuning-operator:tuned" cannot get resource "nodes" in API group "" at the cluster scope I0116 18:16:29.348576 7008 openshift-tuned.go:883] Increasing resyncPeriod to 102 I0116 18:18:11.348742 7008 openshift-tuned.go:209] Extracting tuned profiles I0116 18:18:11.350230 7008 openshift-tuned.go:739] Resync period to pull node/pod labels: 102 [s] E0116 18:18:11.355820 7008 openshift-tuned.go:881] Error getting node ip-10-0-48-132.us-east-2.compute.internal: nodes "ip-10-0-48-132.us-east-2.compute.internal" is forbidden: User "system:serviceaccount:openshift-cluster-node-tuning-operator:tuned" cannot get resource "nodes" in API group "" at the cluster scope I0116 18:18:11.355836 7008 openshift-tuned.go:883] Increasing resyncPeriod to 204 I0116 18:21:35.356153 7008 openshift-tuned.go:209] Extracting tuned profiles
Proposed upstream fix: https://github.com/openshift/cluster-node-tuning-operator/pull/103
Verification positive! After upgrade from 4.3.0-0.nightly-2020-01-26-134543 -> 4.4.0-0.nightly-2020-01-24-141203 there is no more problems with getting 'node' resources.
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, 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-2020:0581