Kubernetes CS

Kubectl of #

Keywoards #

• pods
• ns: namespace(s)
• sc: storageclass(s)
• pv: persistentvolumes
• pvc: persistentvolumeclaim
• deploy: deployment(s)

Actions #

• get
• describe
• delete
• explain

Cheatsheet #

kubectl get pods -A

Resources #

kubectl get resource-type # pods,  ns etc

kubectl describe crd elasticsearch

Services #

Create manifest yaml from currently running service #

kubectl get svc wso2apim-service -o yaml

Endpoints #

kubectl get endpoints

Delete #

Delete namespace #

kubectl delete ns elastic-system

Delete all resources of a name space #

kubectl delete all --all -n default

Delete deployment #

kubectl delete deploy name # kubectl get deploy

Managing node #

ref

Mark node as unschedulable #

kubectl cordon NODE

Mark node as schedulable #

kubectl uncordon NODE

Evict pods #

kubectl drain NODE
# to undo
kubectl uncordon NODE

Procedures #

To delete a pod or deployment #

1. cordon: makes nodes unschedulable
2. drain: deletes pods
3. kubectl delete -f manifest.yaml

Debugging #

Exec into the container #

kubectl exec -it name-of-the-pod -- /bin/sh (command name)

Logs #

kubectl logs -f pod-name
kubectl -n elastic-system logs -f statefulset.apps/elastic-operator # name of operator

Rollout status #

kubectl rollout status deployment deployment-name # kubectl get deployments

Logs of pod before crashing #

kubectl logs -f quickstart-es-default-0 -p #-p does the trick

import os
print(os.env)

Run busybox image temporarily #

the pod will be removed once you are out of the pod shell.

kubectl run -it --rm --restart=Never busybox --image=gcr.io/google-containers/busybox sh

Run a command inside a pod #

kubectl exec shell-demo -- ps aux
kubectl exec shell-demo -- ls /
kubectl exec shell-demo -- cat /proc/1/mounts

Frequently used #

1. Service Port forwarding

kubectl port-forward svc/my-release-prometheus-server 8088:80 # host:service port