How to Deploy Your First Pod in Kubernetes

A Pod is Kubernetes' atomic unit — one or more containers sharing a network namespace and storage. You almost never deploy a bare Pod in production (Deployments and StatefulSets do that for you), but understanding Pods directly is the foundation.

Prerequisites

• A running cluster — k3s, kind, minikube, or full kubeadm. See Install k3s for the easy path.
• kubectl configured (kubectl get nodes works).

Step 1: Imperative first — get something running

kubectl run nginx --image=nginx:1.27-alpine --restart=Never
kubectl get pods
kubectl get pod nginx -o wide

--restart=Never makes it a bare Pod. Without that flag, kubectl run would create a Deployment.

Watch it come up:

kubectl wait --for=condition=Ready pod/nginx --timeout=60s
kubectl describe pod nginx | tail -30

Step 2: Declarative — write the YAML

pod.yaml:

apiVersion: v1
kind: Pod
metadata:
  name: web
  labels:
    app: web
spec:
  containers:
    - name: nginx
      image: nginx:1.27-alpine
      ports:
        - containerPort: 80
      resources:
        requests:
          cpu: 50m
          memory: 64Mi
        limits:
          cpu: 200m
          memory: 128Mi
      readinessProbe:
        httpGet:
          path: /
          port: 80
        initialDelaySeconds: 2
        periodSeconds: 5
      livenessProbe:
        httpGet:
          path: /
          port: 80
        initialDelaySeconds: 10
        periodSeconds: 10

Apply and inspect:

kubectl apply -f pod.yaml
kubectl get pods -o wide
kubectl describe pod web | tail -20

Step 3: Reach the pod

# port-forward from your workstation
kubectl port-forward pod/web 8080:80 &
curl http://localhost:8080/

# exec a shell inside
kubectl exec -it web -- sh

# tail logs
kubectl logs web -f

Step 4: Init containers and sidecars

For pods that need setup before the main container runs:

spec:
  initContainers:
    - name: wait-for-db
      image: busybox:1.36
      command: ['sh', '-c', 'until nc -z db 5432; do sleep 2; done']
  containers:
    - name: app
      image: skyline/api:1.0
    - name: log-shipper       # sidecar
      image: skyline/log-shipper:1.0

The init container must exit 0 before the main containers start.

Step 5: Pod lifecycle and restart policy

Three restartPolicy values:

• Always — default for Deployments
• OnFailure — for Job-like workloads
• Never — for one-shot diagnostic pods

spec:
  restartPolicy: OnFailure
  containers:
    - name: backup
      image: skyline/db-backup:1.0

Watch state transitions:

kubectl get pods -w

States you will see: Pending → ContainerCreating → Running → Succeeded/Failed/CrashLoopBackOff.

Step 6: Delete and clean up

kubectl delete pod web
kubectl delete -f pod.yaml

Pods do not respawn — they are gone forever. Use Deployments in production. A bare Pod is only for one-off diagnostics or learning.

Verify

kubectl get pods
kubectl get events --sort-by=.lastTimestamp | tail -20
kubectl top pod                # if metrics-server is installed

Conclusion

A Pod is the smallest unit you deploy. Knowing how to write one (and read one's events / logs / describe output) is the foundation for understanding Deployments, StatefulSets, DaemonSets, and Jobs — which all just template Pods.

Next steps

• Expose this pod to the network via Kubernetes service with NodePort.
• Set up a cluster first if you skipped ahead — k3s lightweight Kubernetes.
• For container basics see Docker + Compose on Ubuntu.