Skip to main content
Deploying your Docker image to Google Kubernetes Engine (GKE) can be fully automated using Cloud Build. In this guide, we’ll update our CI/CD pipeline by extending cloudbuild.yaml to:
  1. Build the Docker image
  2. Push it to Google Container Registry (GCR)
  3. Deploy to a GKE cluster with a Kubernetes manifest

​
Prerequisites

  • A GKE cluster up and running
  • gcloud CLI configured with your project and zone
  • Cloud Build API enabled
  • Service account with the following roles:
    • roles/container.developer
    • roles/storage.admin
Ensure your Cloud Build service account has the Kubernetes Engine Developer and Storage Admin roles. Without these, build or deploy steps may fail.

​
Kubernetes Deployment Manifest

We’ve already created a Kubernetes Deployment manifest (gke.yaml) for our sample app: 
apiVersion: apps/v1
kind: Deployment
metadata:
  name: gcp-devops-gke
spec:
  replicas: 1
  selector:
    matchLabels:
      app: gcp
  template:
    metadata:
      labels:
        app: gcp
    spec:
      containers:
        - name: gcp-devops-gke
          image: gcr.io/kodekloud-gcp-training/test-gcpdevops:latest
          ports:
            - containerPort: 5000
          env:
            - name: PORT
              value: "5000"
This Deployment exposes port 5000 and pulls the image from GCR. Next, let’s configure Cloud Build.

​
Initial cloudbuild.yaml

Our starting cloudbuild.yaml only built and pushed the Docker image: 
steps:
  - name: 'gcr.io/cloud-builders/docker'
    args:
      ['build', '-t', 'gcr.io/$PROJECT_ID/gcpdevops', '.']
images:
  - 'gcr.io/$PROJECT_ID/gcpdevops'

​
Extending cloudbuild.yaml for GKE Deployment

We’ll add a third step to invoke the gke-deploy builder, which applies our Kubernetes manifest directly to GKE: 
steps:
  # 1. Build the container image
  - name: 'gcr.io/cloud-builders/docker'
    args:
      ['build', '-t', 'gcr.io/$PROJECT_ID/gcpdevops', '.']

  # 2. Push the container image
  - name: 'gcr.io/cloud-builders/docker'
    args:
      ['push', 'gcr.io/$PROJECT_ID/gcpdevops']

  # 3. Deploy to GKE
  - name: 'gcr.io/cloud-builders/gke-deploy'
    args:
      [
        'run',
        '--filename=gke.yaml',
        '--image=gcr.io/$PROJECT_ID/gcpdevops',
        '--location=us-central1-c',
        '--cluster=gke-gcp-devops',
        '--namespace=gcp-devops-prod'
      ]

images:
  - 'gcr.io/$PROJECT_ID/gcpdevops'

​
Step-by-Step Overview

StepBuilder ImageArgumentsPurpose
1gcr.io/cloud-builders/dockerbuild -t gcr.io/$PROJECT_ID/gcpdevops .Build the container image
2gcr.io/cloud-builders/dockerpush gcr.io/$PROJECT_ID/gcpdevopsPush image to Google Container Registry
3gcr.io/cloud-builders/gke-deployrun --filename=gke.yaml --image=gcr.io/$PROJECT_ID/gcpdevops --location=us-central1-c --cluster=gke-gcp-devops --namespace=gcp-devops-prodDeploy manifest to your GKE cluster

​
References

​
Next Steps

  1. Commit and push your updated cloudbuild.yaml to a feature branch.
  2. Open a pull request for review.
  3. Merge into main.
Once merged, Cloud Build will automatically execute the pipeline, build and push your container, then deploy the updated manifest to your GKE cluster. Finally, use: 
kubectl get pods --namespace gcp-devops-prod
to verify your application is running successfully.

Watch Video