Deployments Update and Rollback

In Kubernetes, Deployments automate application updates, versioning, and rollbacks. This guide covers:

• How rollouts create revisions
• Deployment update strategies
• Applying and inspecting updates
• Undoing changes

Rollouts and Versioning

Whenever you create or modify a Deployment, Kubernetes starts a new rollout, creating a revision:

To monitor your rollout and review history:

kubectl rollout status deployment/myapp-deployment
# Waiting for rollout to finish: 0 of 10 updated replicas are available...
# ...
kubectl rollout history deployment/myapp-deployment
# REVISION  CHANGE-CAUSE
# 1         initial create
# 2         updated image to nginx:1.7.1

Deployment Strategies

Kubernetes supports two primary update strategies:

Recreate

This strategy terminates all existing Pods before creating new ones—resulting in downtime.

RollingUpdate (default)

With RollingUpdate, old Pods are replaced incrementally, ensuring continuous availability.

If you omit a strategy in your Deployment spec, RollingUpdate applies by default.

Applying Updates

You can update a Deployment by editing its YAML or using kubectl set image.

deployment.yaml:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deployment
  labels:
    app: myapp
    tier: frontend
spec:
  replicas: 3
  selector:
    matchLabels:
      app: myapp
      tier: frontend
  template:
    metadata:
      labels:
        app: myapp
        tier: frontend
    spec:
      containers:
        - name: nginx-container
          image: nginx:1.7.0

Apply the manifest:

kubectl apply -f deployment.yaml
# deployment "myapp-deployment" configured

Or update the image directly:

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

Each change creates a new revision and triggers a rollout.

Inspecting Deployment Details

To see strategy settings, ReplicaSet events, and scaling details:

kubectl describe deployment myapp-deployment

Example for Recreate:

Name:                   myapp-deployment
StrategyType:           Recreate
Replicas:               5 desired | 5 total | 5 available
Events:
  Type    Reason             Age   From                   Message
  ----    ------             ----  ----                   -------
  Normal  ScalingReplicaSet  11m   deployment-controller  Scaled up replica set myapp-deployment-6795… to 5

Example for RollingUpdate:

Name:                   myapp-deployment
StrategyType:           RollingUpdate
RollingUpdateStrategy:  25% max unavailable, 25% max surge
Replicas:               5 desired | 6 total | 4 available | 2 unavailable
Events:
  Type    Reason             Age  From                   Message
  ----    ------             ---- ----                   -------
  Normal  ScalingReplicaSet  1m   deployment-controller  Scaled up replica set myapp-deployment-67c7… to 5

Under the hood, each Deployment manages ReplicaSets:

List ReplicaSets:

kubectl get replicasets
# NAME                         DESIRED CURRENT READY AGE
# myapp-deployment-67c749c58c  0       0       0     22m
# myapp-deployment-75d7bdbd8d  5       5       5     20m

Rolling Back

To revert a faulty rollout:

kubectl rollout undo deployment/myapp-deployment
# deployment "myapp-deployment" rolled back

After rollback, ReplicaSet counts swap:

kubectl get replicasets
# NAME                         DESIRED CURRENT READY AGE
# myapp-deployment-67c749c58c  5       5       5     22m
# myapp-deployment-75d7bdbd8d  0       0       0     20m

Creating a Deployment with kubectl run

Although kubectl run nginx --image=nginx creates a Deployment by default:

kubectl run nginx --image=nginx
# deployment "nginx" created

Summary of Key Commands

References

• Kubernetes Basics
• Deployments
• kubectl Cheat Sheet