How to delete failed, error and CrashLoopBackOff pods in Kubernetes (k8s)

After a series of unexpected power failures and hardware issues, I had a bunch of failed pods in my Kubernetes cluster. How do I delete all failed pods in Kubernetes?

DO THIS WITH CAUTION - This will delete all failed/error pods in your Kubernetes cluster.

Here is a selection of Error and Crash Loop Back Off pods in my Kubernetes cluster.

```bash
$ kubectl get pods
NAME                                                       READY   STATUS             RESTARTS            AGE
alertmanager-kube-prometheus-stack-alertmanager-0          2/2     Running            6 (9m35s ago)       130d
hello-world-6cb7cd95b5-z7qhr                               1/1     Running            7 (9m35s ago)       142d
kube-prometheus-stack-grafana-7664d8545c-k4bmx             0/3     Completed          0                   96d
kube-prometheus-stack-grafana-7664d8545c-k4tgq             3/3     Running            6 (9m36s ago)       96d
kube-prometheus-stack-grafana-7664d8545c-n7zdj             0/3     Completed          3 (127d ago)        130d
kube-prometheus-stack-grafana-7664d8545c-r7xdj             0/3     Completed          0                   99d
kube-prometheus-stack-kube-state-metrics-69d546c6c-j9srj   1/1     Running            3 (9m36s ago)       130d
kube-prometheus-stack-operator-74cf4c448b-8lq44            1/1     Running            3 (9m36s ago)       130d
kube-prometheus-stack-prometheus-node-exporter-27m78       0/1     CrashLoopBackOff   26525 (2m29s ago)   94d
kube-prometheus-stack-prometheus-node-exporter-pm6dk       0/1     CrashLoopBackOff   966 (2m15s ago)     3d9h
kube-prometheus-stack-prometheus-node-exporter-xpd89       0/1     CrashLoopBackOff   309 (2m29s ago)     25h
metrics-server-699ffcdc79-t5lsw                            1/1     Running            6 (9m36s ago)       135d
my-foldingathome-9f7fdb678-bjr6j                           0/1     Error              1                   128d
my-foldingathome-9f7fdb678-qsc57                           1/1     Running            1 (20h ago)         95d
my-squid-76b7d89496-7mzn9                                  1/1     Running            3 (9m36s ago)       46d
nfs-subdir-external-provisioner-66b5dff8cf-xlzgn           1/1     Running            9 (20h ago)         139d
prometheus-kube-prometheus-stack-prometheus-0              2/2     Running            4 (9m36s ago)       3d9h
rabbit-rabbitmq-0                                          1/1     Running            0                   16m
wiki-js-58cb544c76-wr2zc                                   0/1     CrashLoopBackOff   244 (2m7s ago)      139d
wiki-js-postgresql-0                                       1/1     Running            6 (20h ago)         139d

List all failed pods

Run the following command to list all failed pods in your Kubernetes cluster.

$ kubectl get pods -A | grep -v Running

Example output:

[~] $ kubectl get pods -A | grep -v Running
NAMESPACE        NAME                                                       READY   STATUS             RESTARTS            AGE
default          kube-prometheus-stack-grafana-7664d8545c-k4bmx             0/3     Completed          0                   96d
default          kube-prometheus-stack-grafana-7664d8545c-n7zdj             0/3     Completed          3 (127d ago)        130d
default          kube-prometheus-stack-grafana-7664d8545c-r7xdj             0/3     Completed          0                   99d
default          kube-prometheus-stack-prometheus-node-exporter-27m78       0/1     CrashLoopBackOff   26525 (3m30s ago)   94d
default          kube-prometheus-stack-prometheus-node-exporter-pm6dk       0/1     CrashLoopBackOff   966 (3m16s ago)     3d9h
default          kube-prometheus-stack-prometheus-node-exporter-xpd89       0/1     CrashLoopBackOff   309 (3m30s ago)     25h
default          my-foldingathome-9f7fdb678-bjr6j                           0/1     Error              1                   128d

Double check the list

To get list of pods that are in Error or CrashLoopBackOff state, run the following command.

kubectl get pods --all-namespaces | grep -v Running |awk '{print $2 " --namespace=" $1}' | grep -v ^NAME

Example output:

[~] $ kubectl get pods --all-namespaces | grep -v Running |awk '{print $2 " --namespace=" $1}' | grep -v ^NAME
kube-prometheus-stack-grafana-7664d8545c-k4bmx --namespace=default
kube-prometheus-stack-grafana-7664d8545c-n7zdj --namespace=default
kube-prometheus-stack-grafana-7664d8545c-r7xdj --namespace=default
kube-prometheus-stack-prometheus-node-exporter-27m78 --namespace=default
kube-prometheus-stack-prometheus-node-exporter-pm6dk --namespace=default
kube-prometheus-stack-prometheus-node-exporter-xpd89 --namespace=default

Delete all failed pods

Once you are happy with the list of failed pods, you can delete them using the following command.

kubectl get pods --all-namespaces | grep -v Running |awk '{print $2 " --namespace=" $1}' | grep -v ^NAME | xargs kubectl delete pod

Example output:

[~] $ kubectl get pods --all-namespaces | grep -v Running |awk '{print $2 " --namespace=" $1}' | grep -v ^NAME | xargs kubectl delete pod
pod "kube-prometheus-stack-grafana-7664d8545c-k4bmx" deleted
pod "kube-prometheus-stack-grafana-7664d8545c-n7zdj" deleted
pod "kube-prometheus-stack-grafana-7664d8545c-r7xdj" deleted
pod "kube-prometheus-stack-prometheus-node-exporter-27m78" deleted
pod "kube-prometheus-stack-prometheus-node-exporter-pm6dk" deleted
pod "kube-prometheus-stack-prometheus-node-exporter-xpd89" deleted