Deployment

Gyrinx is deployed on Google Cloud Platform using Cloud Run for the application and Cloud SQL for the database. The deployment process is fully automated through GitHub and Google Cloud Build.

Architecture Overview

Deployment Pipeline

Automatic Deployment

1. Code Push - Developer pushes to main branch

2. GitHub Actions - Runs tests and checks

3. Cloud Build Trigger - Automatically triggered on main branch changes

4. Container Build - Docker image built from Dockerfile

5. Deploy to Cloud Run - New image deployed automatically

6. Database Migrations - Run automatically on container startup

Build Configuration

cloudbuild.yaml

steps:
  # Build Docker image
  - name: 'gcr.io/cloud-builders/docker'
    args: ['build', '-t', 'gcr.io/$PROJECT_ID/gyrinx:$COMMIT_SHA', '.']

  # Push to Container Registry
  - name: 'gcr.io/cloud-builders/docker'
    args: ['push', 'gcr.io/$PROJECT_ID/gyrinx:$COMMIT_SHA']

  # Deploy to Cloud Run
  - name: 'gcr.io/cloud-builders/gcloud'
    args: ['run', 'deploy', 'gyrinx',
           '--image', 'gcr.io/$PROJECT_ID/gyrinx:$COMMIT_SHA',
           '--region', 'europe-west2',
           '--platform', 'managed']

Environment Configuration

Production Settings

• Database: Cloud SQL PostgreSQL

• Static Files: Served by WhiteNoise

• Security: HTTPS enforced, secure headers

• Scaling: Automatic based on traffic

Environment Variables

# Database configuration
DB_CONFIG='{"user": "username", "password": "password"}'
DB_NAME=gyrinx
DB_HOST=cloud_sql_proxy_ip
DB_PORT=5432

# Django settings
SECRET_KEY=production_secret_key
ALLOWED_HOSTS='["gyrinx.app", "www.gyrinx.app"]'
CSRF_TRUSTED_ORIGINS='["https://gyrinx.app"]'

# External services
GOOGLE_ANALYTICS_ID=GA_tracking_id
PATREON_HOOK_SECRET=patreon_webhook_secret

Database Management

Migrations

Migrations run automatically when the container starts:

# In Dockerfile
ENTRYPOINT ["./docker/entrypoint.sh"]

# docker/entrypoint.sh
#!/bin/bash
python manage.py migrate --noinput
python manage.py collectstatic --noinput
exec "$@"

Database Backups

• Automatic Backups: Cloud SQL provides automatic daily backups

• Point-in-time Recovery: Available for disaster recovery

• Manual Backups: Can be triggered before major deployments

Database Access

# Connect to production database (requires Cloud SQL Proxy)
gcloud sql connect gyrinx-app-bootstrap-db --user=postgres

# Run migrations manually (if needed)
gcloud run jobs execute migrate-job --region=europe-west2

Monitoring and Observability

Cloud Run Monitoring

• Metrics: Request count, latency, memory usage

• Logs: Application logs and access logs

• Alerts: Configured for high error rates and latency

Health Checks

# In Django urls.py
urlpatterns = [
    path('health/', health_check_view, name='health'),
]

def health_check_view(request):
    # Basic health check
    return JsonResponse({'status': 'healthy'})

Alerting

• Discord Integration: Alerts sent to #ops channel

• Error Tracking: 500 errors and exceptions

• Performance Monitoring: Slow requests and high memory usage

Security

HTTPS and SSL

• Automatic SSL: Cloud Run provides automatic SSL certificates

• Security Headers: Configured in Django settings

• CSRF Protection: Enforced for all POST requests

Access Control

• IAM Roles: Principle of least privilege

• Service Accounts: Separate accounts for different services

• Secret Management: Environment variables for sensitive data

Security Scanning

• Container Scanning: Automatic vulnerability scanning

• Dependency Updates: Regular security updates

• Code Analysis: GitHub Security Advisories

Performance Optimization

Cloud Run Configuration

# service.yaml
spec:
  template:
    metadata:
      annotations:
        autoscaling.knative.dev/maxScale: "10"
        autoscaling.knative.dev/minScale: "0"
        run.googleapis.com/cpu-throttling: "false"
    spec:
      containerConcurrency: 80
      timeoutSeconds: 900
      containers:
      - resources:
          limits:
            cpu: "2"
            memory: "2Gi"

Database Performance

• Connection Pooling: Configured in Django settings

• Read Replicas: Available for read-heavy workloads

• Query Optimization: Regular performance monitoring

Disaster Recovery

Backup Strategy

1. Database Backups: Daily automatic backups with 7-day retention

2. Point-in-time Recovery: Up to 7 days

3. Code Repository: Git provides complete history

4. Container Images: Stored in Container Registry

Recovery Procedures

# Restore from backup (example)
gcloud sql backups restore BACKUP_ID --restore-instance=gyrinx-restored

# Rollback deployment
gcloud run deploy gyrinx --image=gcr.io/PROJECT/gyrinx:PREVIOUS_SHA

Deployment Checklist

Pre-deployment

• Tests pass in CI/CD

• Database migrations tested locally

• Feature flags configured (if applicable)

• Monitoring alerts updated

Post-deployment

• Health check passes

• Key user flows tested

• Error rates monitored

• Performance metrics checked

Troubleshooting

Common Issues

Migration Failures

# Check migration status
python manage.py showmigrations

# Apply specific migration
python manage.py migrate app_name migration_name

Container Start Failures

# Check Cloud Run logs
gcloud logging read "resource.type=cloud_run_revision" --limit=50

# Debug locally
docker run -it --rm gyrinx:latest /bin/bash

Database Connection Issues

# Test database connection
python manage.py dbshell

# Check Cloud SQL status
gcloud sql instances describe gyrinx-app-bootstrap-db

Performance Issues

• Monitor Cloud Run metrics for CPU/memory usage

• Check database query performance

• Review Django debug toolbar output locally

Cost Optimization

Resource Management

• Auto-scaling: Scales to zero when not in use

• Right-sizing: Monitor resource usage and adjust limits

• Database Sizing: Regular review of database performance tiers

Monitoring Costs

• Budget Alerts: Set up billing alerts

• Resource Usage: Regular review of Cloud Console metrics

• Optimization Opportunities: Identify unused resources

Future Improvements

Staging Environment

• Consider adding a staging environment for testing

• Could use Cloud Run with separate database

• Automated testing on staging before production

Advanced Monitoring

• Application Performance Monitoring (APM)

• Real User Monitoring (RUM)

• Error tracking with detailed stack traces

CI/CD Enhancements

• Canary deployments

• Blue-green deployments

• Automated rollback on failures