Rolling Updates and Rollbacks

This lesson explains how Kubernetes manages Deployment rollouts and rollbacks. When you create or update a Deployment, Kubernetes records each rollout as a new revision and creates a corresponding ReplicaSet. Revisioning makes it easy to track changes and revert to a previous working state when needed.

A slide titled "Rollout and Versioning" showing two revision groups of container icons: Revision 1 with multiple containers labeled nginx:1.7.0 and Revision 2 with containers labeled nginx:1.7.1, plus a central update/rollout icon.

Check rollout progress and history

Check rollout status:

kubectl rollout status deployment/myapp-deployment

Example output:

$ kubectl rollout status deployment/myapp-deployment
Waiting for rollout to finish: 0 of 10 updated replicas are available...
Waiting for rollout to finish: 1 of 10 updated replicas are available...
Waiting for rollout to finish: 2 of 10 updated replicas are available...
Waiting for rollout to finish: 3 of 10 updated replicas are available...
Waiting for rollout to finish: 4 of 10 updated replicas are available...
Waiting for rollout to finish: 5 of 10 updated replicas are available...
Waiting for rollout to finish: 6 of 10 updated replicas are available...
Waiting for rollout to finish: 7 of 10 updated replicas are available...
Waiting for rollout to finish: 8 of 10 updated replicas are available...
Waiting for rollout to finish: 9 of 10 updated replicas are available...
deployment "myapp-deployment" successfully rolled out

View rollout history and revisions:

kubectl rollout history deployment/myapp-deployment

Example output:

$ kubectl rollout history deployment/myapp-deployment
deployments "myapp-deployment"
REVISION  CHANGE-CAUSE
1         <none>
2         kubectl apply --filename=deployment-definition.yml --record=true

The CHANGE-CAUSE column appears only if a change-cause was recorded during the update (historically via --record). That flag has been deprecated in newer kubectl versions and may not be set.

Deployment update strategies

There are two common update strategies for Deployments:

Recreate: Kubernetes scales down the old ReplicaSet to zero replicas, then scales up the new ReplicaSet. This approach can cause downtime during the transition.
RollingUpdate (default): Kubernetes incrementally replaces old Pods with new Pods so that the application stays available during the upgrade.

If no strategy is specified in the Deployment manifest, Kubernetes uses RollingUpdate by default.

By default, Deployments use the RollingUpdate strategy to avoid downtime during upgrades.

A diagram titled "Deployment Strategy" comparing Recreate and Rolling Update, showing rows of nginx container icons labeled nginx:1.7.0 and nginx:1.7.1. Red downward arrows indicate old pods being stopped and green upward arrows show new pods starting, with an "Application down" box in the Recreate flow.

How to update a Deployment

“Updating” a Deployment means changing the pod template: for example, updating the container image, changing labels, or modifying replica counts. Two common approaches:

Update the manifest and apply:

kubectl apply -f deployment-definition.yml

This updates the stored manifest and triggers a new rollout.

Change the image directly:

kubectl set image deployment/myapp-deployment \
  nginx-container=nginx:1.9.1

Note: Using kubectl set image updates the live Deployment in the cluster but does not update your local manifest file — this can cause drift between your repository and cluster state.

If you use kubectl set image or other imperative commands, remember to update your source manifest to avoid drift between your Git/manifest and the cluster.

Example Deployment manifest (corrected YAML):

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deployment
  labels:
    app: myapp
    type: front-end
spec:
  replicas: 3
  selector:
    matchLabels:
      type: front-end
  template:
    metadata:
      name: myapp-pod
      labels:
        app: myapp
        type: front-end
    spec:
      containers:
      - name: nginx-container
        image: nginx:1.7.1

Example commands and output:

$ kubectl apply -f deployment-definition.yml
deployment "myapp-deployment" configured

$ kubectl set image deployment/myapp-deployment \
    nginx-container=nginx:1.9.1
deployment "myapp-deployment" image updated

Inspecting Deployment details

Use kubectl describe deployment to inspect a Deployment’s strategy, replica counts, annotations, and controller events. Example kubectl describe output for a Recreate strategy:

$ kubectl describe deployment myapp-deployment
Name:                   myapp-deployment
Namespace:              default
CreationTimestamp:      Sat, 03 Mar 2018 17:01:55 +0800
Labels:                 app=myapp
                        type=front-end
Annotations:            deployment.kubernetes.io/revision=2
                        kubectl.kubernetes.io/last-applied-configuration=`<truncated-json>`
                        kubernetes.io/change-cause=kubectl apply --filename=d:\Mumshad Files\...
Selector:               type=front-end
Replicas:               5 desired | 5 updated | 5 total | 5 available | 0 unavailable
StrategyType:           Recreate
MinReadySeconds:        0
Pod Template:
  Labels:               app=myapp
                        type=front-end

  Containers:
   nginx-container:
    Image:              nginx:1.7.1
    Port:               `none`
    Environment:        `none`
    Mounts:             `none`
  Volumes:              `none`
Conditions:
Type           Status  Reason
----           ------  ------
Available      True    MinimumReplicasAvailable
Progressing    True    NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet:  myapp-deployment-54c7d6ccc (5/5 replicas created)
Events:
  Type    Reason               Age   From                  Message
  Normal  ScalingReplicaSet    11m   deployment-controller  Scaled up replica set myapp-deployment-6795844b58 to 5
  Normal  ScalingReplicaSet    1m    deployment-controller  Scaled down replica set myapp-deployment-6795844b58 to 0
  Normal  ScalingReplicaSet    56s   deployment-controller  Scaled up replica set myapp-deployment-54c7d6ccc to 5

Example kubectl describe output for a RollingUpdate strategy:

$ kubectl describe deployment myapp-deployment
Name:                   myapp-deployment
Namespace:              default
CreationTimestamp:      Sat, 03 Mar 2018 17:16:53 +0800
Labels:                 app=myapp
                        type=front-end
Annotations:            deployment.kubernetes.io/revision=2
                        kubectl.kubernetes.io/last-applied-configuration=`<truncated-json>`
                        kubernetes.io/change-cause=kubectl apply --filename=d:\Mumshad Files\...
Selector:               type=front-end
Replicas:               5 desired | 5 updated | 6 total | 4 available | 2 unavailable
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge
Pod Template:
  Labels:               app=myapp
                        type=front-end

  Containers:
   nginx-container:
    Image:              nginx
    Port:               `none`
    Environment:        `none`
    Mounts:             `none`
  Volumes:              `none`
Conditions:
Type           Status  Reason
----           ------  ------
Available      True    MinimumReplicasAvailable
Progressing    True    ReplicaSetUpdated
OldReplicaSets: myapp-deployment-67c749c58c (1/1 replicas created)
NewReplicaSet:  myapp-deployment-7d57dbdb8d (5/5 replicas created)
Events:
  Type    Reason               Age  From                  Message
  Normal  ScalingReplicaSet    1m   deployment-controller  Scaled up replica set myapp-deployment-67c749c58c to 5
  Normal  ScalingReplicaSet    1s   deployment-controller  Scaled up replica set myapp-deployment-7d57dbdb8d to 2
  Normal  ScalingReplicaSet    1s   deployment-controller  Scaled down replica set myapp-deployment-67c749c58c to 4
  Normal  ScalingReplicaSet    1s   deployment-controller  Scaled up replica set myapp-deployment-7d57dbdb8d to 3
  Normal  ScalingReplicaSet    0s   deployment-controller  Scaled down replica set myapp-deployment-67c749c58c to 3
  Normal  ScalingReplicaSet    0s   deployment-controller  Scaled up replica set myapp-deployment-7d57dbdb8d to 4
  Normal  ScalingReplicaSet    0s   deployment-controller  Scaled down replica set myapp-deployment-67c749c58c to 2
  Normal  ScalingReplicaSet    0s   deployment-controller  Scaled up replica set myapp-deployment-7d57dbdb8d to 5
  Normal  ScalingReplicaSet    0s   deployment-controller  Scaled down replica set myapp-deployment-67c749c58c to 1

How Deployments upgrade under the hood

Creating a Deployment (for example, with five replicas) creates a ReplicaSet.
The ReplicaSet creates Pods based on the pod template.
When you change the pod template, the Deployment creates a new ReplicaSet and moves Pods from the old ReplicaSet to the new one according to the chosen strategy (Recreate or RollingUpdate). The old ReplicaSet typically remains in the cluster so you can roll back.

A diagram titled "Upgrades" showing a Kubernetes Deployment with two ReplicaSets (ReplicaSet-1 and ReplicaSet-2). Each ReplicaSet contains multiple Pod icons, with some pods visually highlighted to indicate upgrading or replacement.

You can list ReplicaSets to observe the results of upgrades:

kubectl get replicasets

Example before rollback:

$ kubectl get replicasets
NAME                         DESIRED   CURRENT   READY   AGE
myapp-deployment-67c749c58c  0         0         0       22m
myapp-deployment-7d57dbdb8d  5         5         5       20m

Rollback (undo a rollout)

If a new rollout is faulty, you can revert to a previous revision. Undo the most recent rollout:

kubectl rollout undo deployment myapp-deployment

Example:

$ kubectl rollout undo deployment myapp-deployment
deployment "myapp-deployment" rolled back

After the rollback, ReplicaSets reflect the previous revision being scaled back up and the faulty one scaled down:

$ kubectl get replicasets
NAME                         DESIRED   CURRENT   READY   AGE
myapp-deployment-67c749c58c  5         5         5       22m
myapp-deployment-7d57dbdb8d  0         0         0       20m

Quick reference — useful commands

Purpose	Command
Create a Deployment	`kubectl create -f deployment-definition.yml`
List Deployments	`kubectl get deployments`
Apply changes from a manifest	`kubectl apply -f deployment-definition.yml`
Update a Deployment image	`kubectl set image deployment/myapp-deployment nginx-container=nginx:1.9.1`
Check rollout status	`kubectl rollout status deployment/myapp-deployment`
View rollout history	`kubectl rollout history deployment/myapp-deployment`
Undo a rollout (rollback)	`kubectl rollout undo deployment/myapp-deployment`
Inspect Deployment details	`kubectl describe deployment myapp-deployment`
List ReplicaSets	`kubectl get replicasets`

Introduction

Core Concepts

Scheduling

Logging & Monitoring

Application Lifecycle Management

Cluster Maintenance

Security

Storage

Networking

Design and Install a Kubernetes Cluster

Install "Kubernetes the kubeadm way"

Troubleshooting

Helm Basics (2025 Updates)

Kustomize Basics (2025 Updates)

Other Topics

Mock Exams

Rolling Updates and Rollbacks

Check rollout progress and history

Deployment update strategies

How to update a Deployment

Inspecting Deployment details

How Deployments upgrade under the hood

Rollback (undo a rollout)

Quick reference — useful commands

Links and references

Watch Video

Practice Lab

​Check rollout progress and history

​Deployment update strategies

​How to update a Deployment

​Inspecting Deployment details

​How Deployments upgrade under the hood

​Rollback (undo a rollout)

​Quick reference — useful commands

​Links and references

Watch Video

Practice Lab

Check rollout progress and history

Deployment update strategies

How to update a Deployment

Inspecting Deployment details

How Deployments upgrade under the hood

Rollback (undo a rollout)

Quick reference — useful commands

Links and references