How to remove NotReady Pods from Kubernetes

You got a pod left in NotReady status due to failed helm chart installation or resource deployment.

➜  ✗ kubectl get pods
NAME                                                READY   STATUS     RESTARTS   AGE
release-0.11-kube-promethe-admission-create-7j7fz   1/2     NotReady   2          8m1s

Tried to remove the pod with kubectl delete pod will just cause the pod to be recreated.

Run helm ls -A to list all current installed charts. Check the chart status.

NAME            NAMESPACE       REVISION        UPDATED                                 STATUS  CHART                           APP VERSION
tomlab          default         1               2022-09-26 09:30:38.909619 +1000 AEST   failed  kube-prometheus-stack-40.1.2    0.59.1 

To remove the failed chart, run helm uninstall [chart_name] -n [namespace_name]. This should remove the pods created by the chart.

Sometimes a pending deployment will left pods in NotReady status. As long as the deployment is not removed, the pods will get recreated even when you try to delete them manually.

To check currently deployments:

kubectl get deployments --all-namespaces

To remove the deployments and its resources:

kubectl delete deployment [deployment_name]

If all above fails, lookup all the jobs currently pending.

➜ ✗ kubectl get job
NAME                                          COMPLETIONS   DURATION   AGE
release-0.11-kube-promethe-admission-create   0/1           2d17h      2d17h

Kill the job

kubectl delete -f jobs [job_name]

After that you should be able to clean up the NotReady pods relate to the job.