Skip to main content

Deployment Rollout and Rollback

Hands-on example​

Assume:

  • A working Kubernetes cluster
  • kubectl is configured

Initial state​

No Deployments exist.

kubectl get deployments
No resources found in default namespace.

Step 1: Create a Deployment​

kubectl create deployment demo-deploy --image=nginx:1.25
deployment.apps/demo-deploy created

Check status.

kubectl get deployments
NAME          READY   UP-TO-DATE   AVAILABLE   AGE
demo-deploy   1/1     1            1           5s

Check ReplicaSets.

kubectl get rs
NAME                     DESIRED   CURRENT   READY   AGE
demo-deploy-6f7c8d9f6b   1         1         1       5s

What changed:

  • One Deployment and one ReplicaSet exist

What did not change:

  • Only one Pod version is running

Step 2: Update the image (trigger a rollout)​

kubectl set image deployment/demo-deploy nginx=nginx:1.26
deployment.apps/demo-deploy image updated

Watch the rollout.

kubectl rollout status deployment/demo-deploy
deployment "demo-deploy" successfully rolled out

Check ReplicaSets again.

kubectl get rs
NAME                     DESIRED   CURRENT   READY   AGE
demo-deploy-7a8b9c0d1e   1         1         1       10s
demo-deploy-6f7c8d9f6b   0         0         0       1m

What changed:

  • A new ReplicaSet was created
  • Old ReplicaSet was scaled down

What did not change:

  • Deployment name and selector

Step 3: Roll back to the previous version​

kubectl rollout undo deployment/demo-deploy
deployment.apps/demo-deploy rolled back

Verify.

kubectl get rs
NAME                     DESIRED   CURRENT   READY   AGE
demo-deploy-6f7c8d9f6b   1         1         1       2m
demo-deploy-7a8b9c0d1e   0         0         0       1m

What changed:

  • The previous ReplicaSet is active again

What did not change:

  • No Pods were manually deleted

Key observation​

  • Rollouts and rollbacks work only through Deployments
  • Deleting Pods manually bypasses rollout control