KubePodNotReady #

Meaning #

[This alert][KubePodNotReady] is fired when some pods have not been in the Ready state for a certain period. This can have different reasons as described here: When the pod is Running but not Ready, the Readiness probe is failing. An application-specific error may prevent the pod from being attached to a service. When the pod remains in Pending it can not be deployed to particular namespaces and nodes.

Impact #

This pod is not functional and doesn’t receive any traffic. Depending on how many functional replicas are still available, the impact differs.

Diagnosis #

The notification details should list the pod that’s not ready (and the pod’s namespace). E.g.:

 - alertname = KubePodNotReady
...
 - namespace = openshift-logging
 - pod = elasticsearch-cdm-u1gqqbu6-2-868ddd4b45-w224d
...

Start by checking the status of the pod:

$ kubectl get pod -n $NAMESPACE $POD

When the pod state is in Running, you can check its logs:

$ kubectl logs -n $NAMESPACE $POD

Be aware there may be multiple containers in the pod. A check of all their logs may be required. If the pod isn’t running (for instance, if it’s stuck in ContainerCreating), then try to find out why.

Mitigation #

Try to find the issue in the logs before deleting it.