Blue-green deployments in Kubernetes - Commands & Configuration

Introduction

Blue-green deployment is a way to update applications with almost no downtime. It keeps two versions of the app running: one live (blue) and one new (green). You switch traffic to the green version when ready, so users see no interruption.

When you want to update your app without making users wait or see errors.

When you need to test a new app version in production safely before full release.

When you want to quickly roll back to the old version if the new one has problems.

When your app must be available 24/7 without downtime during updates.

When you want to reduce risks of deployment failures by switching traffic smoothly.

Config File - blue-green-deployment.yaml

blue-green-deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-app-blue
  labels:
    app: my-app
    version: blue
spec:
  replicas: 3
  selector:
    matchLabels:
      app: my-app
      version: blue
  template:
    metadata:
      labels:
        app: my-app
        version: blue
    spec:
      containers:
      - name: my-app-container
        image: my-app:blue
        ports:
        - containerPort: 80
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-app-green
  labels:
    app: my-app
    version: green
spec:
  replicas: 3
  selector:
    matchLabels:
      app: my-app
      version: green
  template:
    metadata:
      labels:
        app: my-app
        version: green
    spec:
      containers:
      - name: my-app-container
        image: my-app:green
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: my-app-service
spec:
  selector:
    app: my-app
    version: blue
  ports:
  - protocol: TCP
    port: 80
    targetPort: 80
  type: ClusterIP

This file defines two deployments: my-app-blue and my-app-green. Each runs a different version of the app labeled by version. The Service routes traffic to the blue version initially by selecting pods with label version: blue. To switch traffic, you update the Service selector to version: green.

Commands

This command creates both blue and green deployments and the service routing to blue. It sets up the environment for blue-green deployment.

Terminal

kubectl apply -f blue-green-deployment.yaml

Expected Output

deployment.apps/my-app-blue created

deployment.apps/my-app-green created

service/my-app-service created

Check that pods for both blue and green versions are running correctly before switching traffic.

Terminal

kubectl get pods -l app=my-app -o wide

Expected Output

NAME                          READY   STATUS    RESTARTS   AGE   IP           NODE

my-app-blue-6f7d8c9f7b-abc12  1/1     Running   0          1m    10.244.1.5   node1

my-app-blue-6f7d8c9f7b-def34  1/1     Running   0          1m    10.244.1.6   node2

my-app-green-7g8h9i0j1k-lmn56 1/1     Running   0          1m    10.244.2.7   node3

my-app-green-7g8h9i0j1k-opq78 1/1     Running   0          1m    10.244.2.8   node4

→

-l - Filter pods by label

→

-o wide - Show extra pod details like IP and node

Switch the service selector to route traffic from blue pods to green pods. This changes which version users access without downtime.

Terminal

kubectl patch service my-app-service -p '{"spec":{"selector":{"app":"my-app","version":"green"}}}'

Expected Output

service/my-app-service patched

Verify that the service endpoints now point to green pods after switching traffic.

Terminal

kubectl get endpoints my-app-service

Expected Output

NAME            ENDPOINTS                                   AGE

my-app-service  10.244.2.7:80,10.244.2.8:80                 2m

Key Concept

If you remember nothing else from this pattern, remember: run two app versions side-by-side and switch user traffic by changing the service selector to avoid downtime.

Common Mistakes

Not running the green deployment before switching the service selector

Switching traffic to a non-existent or unhealthy green deployment causes downtime or errors.

Always deploy and verify the green version pods are running and ready before switching the service selector.

Forgetting to update the service selector labels exactly to match the green deployment labels

If labels do not match, the service will have no endpoints and users cannot reach the app.

Double-check the service selector labels match the green deployment pod labels exactly.

Deleting the blue deployment immediately after switching traffic

If the green version has issues, you cannot quickly roll back to blue without the pods running.

Keep the blue deployment running until you confirm the green version works well, then delete blue.

Summary

Create two deployments with different labels for blue and green versions of the app.

Use a Kubernetes Service to route traffic to the blue version initially.

Switch traffic to the green version by patching the service selector to the green labels.

Verify pods and service endpoints before and after switching to ensure smooth deployment.

Practice

1. What is the main purpose of a blue-green deployment in Kubernetes?

easy

A. To update applications without downtime by switching traffic between two versions

B. To create multiple replicas of a pod for load balancing

C. To automatically scale pods based on CPU usage

D. To backup data from one cluster to another

Solution

1. Step 1: Understand blue-green deployment concept

   Blue-green deployment runs two versions of an app side by side to avoid downtime.

2. Step 2: Identify the main goal

   The main goal is to switch user traffic smoothly from the old version (blue) to the new version (green).

3. Final Answer:

   To update applications without downtime by switching traffic between two versions -> Option A

4. Quick Check:

   Blue-green deployment = zero downtime updates [OK]

Hint: Blue-green means two versions, switch traffic smoothly [OK]

Common Mistakes:

• Confusing blue-green with scaling pods
• Thinking it is for backups
• Mixing it with auto-scaling

2. Which Kubernetes resource is typically used to switch traffic between blue and green deployments?

easy

A. ConfigMap

B. PersistentVolume

C. Service

D. Ingress Controller

Solution

1. Step 1: Identify traffic routing resource

   In Kubernetes, a Service routes traffic to pods based on labels.

2. Step 2: Understand blue-green switching

   Switching traffic between blue and green versions is done by changing the Service selector to point to the desired pods.

3. Final Answer:

   Service -> Option C

4. Quick Check:

   Service routes traffic in blue-green deployments [OK]

Hint: Service controls traffic routing between versions [OK]

Common Mistakes:

• Choosing ConfigMap which stores config data
• Selecting PersistentVolume which manages storage
• Picking Ingress Controller which manages external access

3. Given the following Kubernetes Service selector for blue deployment:

selector:
  app: myapp
  version: blue

What happens if you change the selector to version: green?

medium

A. Traffic will be split evenly between blue and green pods

B. Traffic will be routed to pods labeled with version green

C. Traffic will stop because selector is invalid

D. Pods with version blue will receive traffic

Solution

1. Step 1: Understand Service selector role

   The Service selector chooses pods matching the labels to send traffic to.

2. Step 2: Effect of changing selector

   Changing selector to version: green directs traffic only to pods labeled green, ignoring blue pods.

3. Final Answer:

   Traffic will be routed to pods labeled with version green -> Option B

4. Quick Check:

   Selector change = traffic to matching pods [OK]

Hint: Service selector controls which pods get traffic [OK]

Common Mistakes:

• Assuming traffic splits automatically
• Thinking selector change breaks traffic
• Believing old pods still get traffic

4. You deployed a green version but users still see the blue version. What is the most likely cause?

medium

A. The Service selector was not updated to point to green pods

B. The green pods failed to start due to image pull error

C. The Deployment was deleted accidentally

D. The cluster is out of CPU resources

Solution

1. Step 1: Check traffic routing setup

   If users still see blue, traffic is likely still routed to blue pods.

2. Step 2: Identify cause of routing

   This happens if the Service selector was not updated to green pods after deploying green version.

3. Final Answer:

   The Service selector was not updated to point to green pods -> Option A

4. Quick Check:

   Service selector update needed to switch traffic [OK]

Hint: Update Service selector to switch traffic [OK]

Common Mistakes:

• Assuming pods failed without checking
• Thinking Deployment deletion causes this
• Blaming cluster resource issues first

5. You want to perform a blue-green deployment with zero downtime. You have:

• Blue pods running version 1
• Green pods running version 2

Which sequence of steps ensures a safe switch?

hard

A. Update Deployment to green version, wait for rollout, then update Service selector

B. Delete blue pods first, then update Service selector to green pods

C. Scale down blue pods to zero, then create green pods and update Service selector

D. Update Service selector to green pods, then delete blue pods after confirming green is healthy

Solution

1. Step 1: Switch traffic safely

   First, update the Service selector to point to green pods so new traffic goes to version 2.

2. Step 2: Confirm green pods are healthy

   Check green pods are running well before removing blue pods to avoid downtime.

3. Step 3: Remove old version

   After confirmation, delete blue pods to free resources.

4. Final Answer:

   Update Service selector to green pods, then delete blue pods after confirming green is healthy -> Option D

5. Quick Check:

   Switch traffic first, confirm health, then remove old [OK]

Hint: Switch traffic first, confirm green healthy, then remove blue [OK]

Common Mistakes:

• Deleting blue pods before switching traffic
• Not confirming green pods health
• Updating Deployment without switching Service selector