kubectl

kubectl is the command-line interface (CLI) for Kubernetes. It manages resources in the cluster (Pods, Deployments, Services, ConfigMaps, etc.) through the API Server, interacting directly with the Kubernetes API. It reads the ~/.kube/config file to configure the connection to the target cluster. In essence it is a REST client that sends HTTP requests to the Kubernetes API.

Reference docs: kubectl reference (official)

2. Basic Syntax

kubectl [command] [TYPE] [NAME] [flags]

Part	Description
`command`	Operation type, e.g. `get`, `create`, `delete`, `describe`
`TYPE`	Resource type, e.g. `pod`, `deployment`, `service` (abbreviations supported)
`NAME`	Resource name (optional)
`flags`	Parameters, e.g. `-n namespace`, `-o yaml`

✅ Examples:

kubectl get pods -A                   # List all Pods
kubectl describe node node1        # Show details of node1
kubectl delete deployment my-app   # Delete the Deployment named my-app

3. Common Commands Cheat Sheet

Command	Purpose	Example
`kubectl get`	List resources	`kubectl get pods -A`
`kubectl describe`	Show detailed resource info	`kubectl describe pod nginx-123`
`kubectl create`	Create a resource from a file or stdin	`kubectl create -f app.yaml`
`kubectl apply`	Declarative update (recommended)	`kubectl apply -f app.yaml`
`kubectl delete`	Delete a resource	`kubectl delete pod nginx-123`
`kubectl logs`	View container logs	`kubectl logs -f my-pod`
`kubectl exec`	Run a command inside a container	`kubectl exec -it my-pod -- sh`
`kubectl port-forward`	Port-forward (access a service locally)	`kubectl port-forward svc/my-svc 8080:80`
`kubectl proxy`	Start the API proxy	`kubectl proxy --port=8001`
`kubectl top`	Show resource usage (requires Metrics Server)	`kubectl top nodes`
`kubectl rollout`	Manage rollouts / rollback / status	`kubectl rollout undo deploy/my-app`

4. Common Resource Types

Full name	Abbrev.	Description
`pods`	`po`	Smallest scheduling unit
`services`	`svc`	Service discovery & load balancing
`deployments`	`deploy`	Stateless app management
`daemonsets`	`ds`	One Pod per node (e.g. Flannel)
`statefulsets`	`sts`	Stateful apps (e.g. databases)
`configmaps`	`cm`	Non-sensitive config data
`secrets`	`secret`	Sensitive data (base64 encoded)
`namespaces`	`ns`	Namespace (logical isolation)
`nodes`	`no`	Worker node
`persistentvolumes`	`pv`	Storage volume (cluster-scoped)
`persistentvolumeclaims`	`pvc`	Storage claim (namespace-scoped)

✅ Use abbreviations to work faster: kubectl get po instead of kubectl get pods

5. Common Flags

Flag	Description
`-n <namespace>`	Target a namespace (default `default`)
`-A` / `--all-namespaces`	All namespaces
`-o wide`	Show more info (IP, Node, etc.)
`-o yaml` / `-o json`	Output as YAML/JSON
`-l key=value`	Filter resources by label
`-w`	Watch in real time
`--dry-run=client`	Dry run (no actual creation)
`--force`	Force delete (with `--grace-period=0`)

✅ Examples:

# Show IP and node for every Pod in kube-system
kubectl get pods -n kube-system -o wide

# Watch Pod changes in the default namespace in real time
kubectl get pods -w

6. Typical Usage Scenarios

1. Troubleshooting a Pod

kubectl get pods                          # Check status
kubectl describe pod <pod-name>           # Inspect events and config
kubectl logs <pod-name>                   # View logs
kubectl exec -it <pod-name> -- sh         # Enter the container

2. Deploying an Application

kubectl apply -f deployment.yaml          # Deploy
kubectl get deploy                        # Check deployment status
kubectl rollout status deploy/my-app      # Check rollout status

3. Debugging a Service

# Forward local port 8080 to the Service's port 80
kubectl port-forward svc/my-service 8080:80

# Test whether it is reachable
curl http://localhost:8080

4. Checking Cluster Status

kubectl get nodes                         # Are nodes Ready?
kubectl top nodes                         # CPU/memory usage (needs Metrics Server)

7. Config File: `~/.kube/config`

Contains cluster, user, and context information.
Manage it with these commands:

kubectl config view                # View current config
kubectl config use-context prod    # Switch context
kubectl config get-contexts        # List all contexts

This file is critical when managing multiple clusters.

8. Advanced Tips

1. Shell Auto-Completion (Bash/Zsh)

# Bash
echo "source <(kubectl completion bash)" >> ~/.bashrc
source ~/.bashrc

# Zsh
echo "source <(kubectl completion zsh)" >> ~/.zshrc

2. Aliases

alias k=kubectl
alias kg='kubectl get'
alias kd='kubectl describe'

3. Custom Output Columns

kubectl get pods -o custom-columns=NAME:.metadata.name,NODE:.spec.nodeName,STATUS:.status.phase

4. Bulk Deletion

# Delete all Jobs in Completed state
kubectl delete job -l job-name=my-job

# Delete all Pods that are not Running
kubectl delete pod --field-selector=status.phase!=Running

9. Caveats

Access control: Operations are restricted by RBAC; ordinary users may be unable to delete system Pods.
Namespace isolation: When -n is omitted, the default namespace is targeted.
apply vs create:
- apply: declarative, supports incremental updates (recommended).
- create: imperative, re-running it errors out.
Deleting resources: Some resources (e.g. PVC) persist after the Pod is deleted and must be cleaned up manually.

10. Learning Resources

Official docs: https://kubernetes.io/docs/reference/kubectl/
Interactive tutorial: https://kubernetes.io/docs/tutorials/kubernetes-basics/
Cheat sheet: print the kubectl cheatsheet

Summary

kubectl is the “Swiss Army knife” of Kubernetes:

Inspect: get, describe, logs
Operate: apply, delete, exec
Debug: port-forward, proxy, top
Manage: config, rollout, scale

Master kubectl, and you master the lifeblood of Kubernetes!