DEV Community

Cover image for Part-118: 🚀Implement Vertical Pod Autoscaling (VPA) in Google Kubernetes Engine (GKE)
Latchu@DevOps
Latchu@DevOps

Posted on

Part-118: 🚀Implement Vertical Pod Autoscaling (VPA) in Google Kubernetes Engine (GKE)

#kubernetes #devops #googlecloud #devto

Vertical Pod Autoscaling (VPA) in GKE automatically adjusts CPU and memory requests for your workloads, ensuring pods get exactly what they need — no more, no less.

In this guide, we’ll walk through how to implement VPA for a sample app in GKE.

🧩 Step 1: Introduction

We’ll implement Vertical Pod Autoscaling in a GKE cluster.

Pre-requisite:

Make sure Cluster Autoscaler is enabled on your node pool.
This allows GKE to scale nodes automatically when VPA increases resource requests.

📁 Step 2: Review Kubernetes Manifests

Create a new directory for your manifests:

mkdir kube-manifests-vpa
cd kube-manifests-vpa
Enter fullscreen mode Exit fullscreen mode

01-kubernetes-deployment.yaml

apiVersion: apps/v1
kind: Deployment 
metadata: 
  name: myapp1-deployment
spec: 
  replicas: 1
  selector:
    matchLabels:
      app: myapp1
  template:  
    metadata:
      name: myapp1-pod
      labels:
        app: myapp1  
    spec:
      containers: 
        - name: myapp1-container
          image: ghcr.io/stacksimplify/kubenginx:1.0.0
          ports: 
            - containerPort: 80  
          resources:
            requests:
              memory: "5Mi"
              cpu: "25m"
            limits:
              memory: "50Mi"
              cpu: "50m"
Enter fullscreen mode Exit fullscreen mode

02-kubernetes-cip-service.yaml

apiVersion: v1
kind: Service 
metadata:
  name: myapp1-cip-service
spec:
  type: ClusterIP
  selector:
    app: myapp1
  ports: 
    - name: http
      port: 80
      targetPort: 80
Enter fullscreen mode Exit fullscreen mode

🚀 Step 3: Deploy the Kubernetes Resources 

# Deploy
kubectl apply -f kube-manifests-vpa

# Verify
kubectl get deploy
kubectl get pods
Enter fullscreen mode Exit fullscreen mode

v1

⚙️ Step 4: Enable Vertical Pod Autoscaler (VPA)

You can do this either via GKE Console or YAML.

Option 1: Using GKE Console

Go to Workloads → myapp1-deployment
Click Configure
Under Vertical Pod Autoscaling > Configure

  • Mode: Auto
  • Define Resource Policies
  • Add your default container and save

Click Save

Option 2: Using YAML

Create 03-vpa.yaml:

apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: myapp1-deployment
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: myapp1-deployment
  updatePolicy:
    updateMode: Auto
  resourcePolicy:
    containerPolicies:
      - containerName: myapp1-container
        mode: Auto
        controlledResources: ["cpu", "memory"]
        minAllowed:
          cpu: 25m
          memory: 50Mi
        maxAllowed:
          cpu: 100m
          memory: 100Mi
Enter fullscreen mode Exit fullscreen mode

Apply it:

kubectl apply -f 03-vpa.yaml
Enter fullscreen mode Exit fullscreen mode

🧠 Step 5: Check VPA Status 

kubectl get vpa
Enter fullscreen mode Exit fullscreen mode

Example output:

NAME                MODE   CPU   MEM    PROVIDED   AGE
myapp1-deployment   Auto   25m   50Mi   True       6m
Enter fullscreen mode Exit fullscreen mode

Describe for more details:

kubectl describe vpa myapp1-deployment
Enter fullscreen mode Exit fullscreen mode

v2

💥 Step 6: (Optional) Quick Load Test

Run a simple load generator to trigger autoscaling:

kubectl run -i --tty load-generator --rm --image=busybox --restart=Never -- /bin/sh -c "while sleep 0.01; do wget -q -O- http://myapp1-cip-service; done"
Enter fullscreen mode Exit fullscreen mode

Watch your pod being recreated with updated CPU and memory requests as VPA adjusts resources.

🧹 Step 7: Clean Up 

kubectl delete -f kube-manifests-vpa
kubectl delete vpa myapp1-deployment
Enter fullscreen mode Exit fullscreen mode

✅ Summary

  • VPA helps right-size your workloads automatically
  • It works great with Cluster Autoscaler
  • In Autopilot clusters, VPA is enabled by default
  • In Standard clusters, you need to enable VPA per workload

With VPA, your GKE cluster stays efficient — pods use exactly what they need, and you spend less time guessing resource values.

🌟 Thanks for reading! If this post added value, a like ❤️, follow, or share would encourage me to keep creating more content.

— Latchu | Senior DevOps & Cloud Engineer

☁️ AWS | GCP | ☸️ Kubernetes | 🔐 Security | ⚡ Automation
📌 Sharing hands-on guides, best practices & real-world cloud solutions

Top comments (0)