Skip to content

Automatic Deployment Setup

Overview

Your GitHub Actions workflow now automatically deploys to Kubernetes environments after building and pushing Docker images. This document explains the required setup and secrets.

Automatic Deployment Flow

Branch-to-Environment Mapping

  • main branch → Production environment (prod namespace)
  • staging branch → Staging environment (staging namespace)
  • develop branch → Development environment (dev namespace)

What Happens Automatically

  1. Code is pushed to main, staging, or develop branch
  2. Docker image is built and pushed to Google Artifact Registry
  3. Deployment is automatically restarted in the corresponding environment
  4. Workflow waits for rollout to complete (5 minute timeout)
  5. Pod status is verified and reported

Required GitHub Secrets

You need to configure these secrets in your GitHub repository settings:

Existing Secrets

  • GCP_SERVICE_ACCOUNT_KEY - Already configured for Docker builds
  • DATABASE_URL - Already configured for builds

Configuration

The workflow uses the following hardcoded values for the GKE cluster:

  • Cluster Name: kame-house
  • Cluster Region: us-central1 (regional cluster)
  • Project: kame-457417

No additional secrets are required for deployment since cluster details are not sensitive.

Setup Requirements

Only the existing secrets are needed:

  • GCP_SERVICE_ACCOUNT_KEY: Already configured for Docker builds and deployments
  • DATABASE_URL: Already configured for builds

Service Account Permissions

Ensure your service account (GCP_SERVICE_ACCOUNT_KEY) has these IAM roles:

  • roles/container.developer - To access GKE clusters
  • roles/artifactregistry.writer - To push Docker images (already configured)

Workflow Behavior

Successful Deployment

  • ✅ Image built and pushed
  • ✅ Deployment restarted
  • ✅ Rollout completed successfully
  • ✅ Pod status verified

Branch Protection

  • Only main, staging, and develop trigger deployments
  • Pull requests build images but don't deploy
  • Other branches build images but don't deploy

Timeout and Error Handling

  • Deployment rollout timeout: 5 minutes
  • Failed deployments will fail the workflow
  • Successful builds with failed deployments will show which step failed

Verification Commands

After setup, you can verify deployments:

# Check deployment status
kubectl rollout status deployment/portfolio -n prod
kubectl rollout status deployment/portfolio -n staging
kubectl rollout status deployment/portfolio -n dev

# View pods
kubectl get pods -n prod -l app=portfolio
kubectl get pods -n staging -l app=portfolio
kubectl get pods -n dev -l app=portfolio

Benefits

✅ Zero-click deployments - Push code, get deployed automatically ✅ Environment parity - Same process for all environments
✅ Fast updates - No waiting for manual deployment ✅ Visibility - Deployment status in GitHub Actions ✅ Rollback ready - Standard Kubernetes rollout capabilities

Rollback Process

If you need to rollback a deployment:

# Rollback to previous version
kubectl rollout undo deployment/portfolio -n prod
kubectl rollout undo deployment/portfolio -n staging
kubectl rollout undo deployment/portfolio -n dev

# Rollback to specific revision
kubectl rollout undo deployment/portfolio --to-revision=2 -n prod

Monitoring

  • GitHub Actions will show deployment status
  • Kubernetes events will be logged
  • Failed deployments will fail the workflow
  • Successful deployments will be confirmed in the workflow summary