etcdDatabaseHighFragmentationRatio

Find your pod

oc get pods -n openshift-etcd | grep etcd

Connect to the pod

oc rsh -n openshift-etcd <pod name>

Operation

unset ETCDCTL_ENDPOINTS
etcdctl --command-timeout=30s --endpoints=https://localhost:2379 defrag

SamplesMissingTBRCredential

According to https://github.com/openshift/cluster-samples-operator/issues/404.

They have “la flemme”.

InsightsDisabled

Due to common code with RedHat Openshift, OKD displays this info message, you can silent this message.

Or (not recommended in open-source), you can enable remote insights: https://docs.openshift.com/container-platform/4.12/support/remote_health_monitoring/enabling-remote-health-reporting.html

AlertmanagerReceiversNotConfigured

You can ignore if you want, but it’s better configured.

operator-lifecycle-manager: Cannot update - Waiting for updates to take effect

You need to reboot your nodes.

sudo shutdown -r now

Unable to restart cluster after a long shutdown

As described on this page, a long shutdown will not renew certificates.

• ssh core@[node ip]

• sudo -i to pass as root

• export KUBECONFIG=/etc/kubernetes/static-pod-resources/kube-apiserver-certs/secrets/node-kubeconfigs/lb-int.kubeconfig

• oc get csr -o name | xargs oc adm certificate approve

Wait a bit (a lot of time) to start the cluster