Continuous Deployment¶

Strategies for automated deployments.

Deployment Strategies¶

1. Trunk-Based Development¶

Deploy on every push to main:

# .github/workflows/deploy.yaml
on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: go install github.com/plexusone/omnideploy/cmd/omnideploy@latest
      - run: omnideploy up --config deploy.yaml --yes

2. Release-Based¶

Deploy only on version tags:

on:
  push:
    tags: ['v*']

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: omnideploy up --config deploy.yaml --stack production --yes

3. Environment Promotion¶

main → staging → production

on:
  push:
    branches: [main]

jobs:
  deploy-staging:
    runs-on: ubuntu-latest
    environment: staging
    steps:
      - run: omnideploy up --config deploy.yaml --stack staging --yes

  deploy-production:
    needs: deploy-staging
    runs-on: ubuntu-latest
    environment: production
    steps:
      - run: omnideploy up --config deploy.yaml --stack production --yes

Rollback Strategies¶

Manual Rollback¶

# Deploy previous version
omnideploy up --config deploy.yaml --stack production --yes
# (with previous image tag in config)

Automated Rollback¶

deploy:
  runs-on: ubuntu-latest
  steps:
    - name: Deploy
      id: deploy
      run: omnideploy up --config deploy.yaml --yes

    - name: Health check
      id: health
      run: curl -f https://my-app.example.com/health

    - name: Rollback on failure
      if: failure() && steps.deploy.outcome == 'success'
      run: |
        git checkout HEAD~1 -- deploy.yaml
        omnideploy up --config deploy.yaml --yes

Blue-Green Deployments¶

Deploy new version alongside old, then switch:

# Deploy new version to blue stack
omnideploy up --config deploy.yaml --stack blue --yes

# Test blue
curl https://blue.example.com/health

# Switch traffic (DNS or load balancer)
# ...

# Destroy old green
omnideploy destroy --stack green --yes

Canary Deployments¶

Not directly supported by LightSail. Use ECS or Kubernetes targets for canary deployments with traffic splitting.

Monitoring Deployments¶

Deployment Notifications¶

- name: Notify Slack
  if: always()
  uses: slackapi/slack-github-action@v1
  with:
    payload: |
      {
        "text": "Deployment ${{ job.status }}: ${{ github.repository }}"
      }
  env:
    SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK }}

Health Checks¶

- name: Verify deployment
  run: |
    for i in {1..30}; do
      if curl -sf https://my-app.example.com/health; then
        echo "Deployment healthy"
        exit 0
      fi
      sleep 10
    done
    echo "Health check failed"
    exit 1

Best Practices¶

Always preview first in CI before deploying
Use environments for approval gates
Monitor after deploy with health checks
Have rollback plan ready
Keep deployments small and frequent
Use semantic versioning for releases