GCP Cloud Run Deployment¶

Deploy Schopenhauer as a serverless API on Google Cloud Run.

Prerequisites¶

Google Cloud SDK: Install and configure gcloud CLI
GCP Project: A GCP project with billing enabled
APIs Enabled: Cloud Run API, Artifact Registry API

Quick Deployment¶

Use the included deployment script:

# Basic deployment
./deploy.sh --project your-project-id

# Custom region
./deploy.sh --project your-project-id --region us-west1

# Custom resources
./deploy.sh --project your-project-id \
    --memory 1Gi \
    --cpu 2 \
    --max-instances 20

Manual Deployment¶

1. Build the Container¶

# Set your project
gcloud config set project YOUR_PROJECT_ID

# Build and push image
gcloud builds submit --tag gcr.io/YOUR_PROJECT_ID/schopenhauer

2. Deploy to Cloud Run¶

gcloud run deploy schopenhauer-api \
    --image gcr.io/YOUR_PROJECT_ID/schopenhauer \
    --region us-central1 \
    --platform managed \
    --memory 512Mi \
    --cpu 1 \
    --min-instances 0 \
    --max-instances 10 \
    --concurrency 80 \
    --timeout 300s \
    --allow-unauthenticated

3. Get Service URL¶

gcloud run services describe schopenhauer-api \
    --region us-central1 \
    --format 'value(status.url)'

Configuration Options¶

Resource Allocation¶

Setting	Default	Description
`--memory`	512Mi	Memory allocation
`--cpu`	1	CPU allocation
`--min-instances`	0	Minimum instances (0 = scale to zero)
`--max-instances`	10	Maximum instances
`--concurrency`	80	Max concurrent requests per instance
`--timeout`	300s	Request timeout

Environment Variables¶

Set environment variables in Cloud Run:

gcloud run deploy schopenhauer-api \
    --set-env-vars "ENVIRONMENT=production,LOG_LEVEL=info"

Using the Deployed API¶

Health Check¶

curl https://YOUR-SERVICE-URL/health

Generate Document¶

curl -X POST "https://YOUR-SERVICE-URL/generate" \
    -H "Content-Type: application/json" \
    -d '{
      "title": "My Report",
      "sections": [
        {"type": "heading", "title": "Hello", "level": 1}
      ]
    }' \
    --output report.docx

Using CLI¶

will cloud health --url https://YOUR-SERVICE-URL
will cloud generate spec.yaml -o doc.docx --url https://YOUR-SERVICE-URL

Authentication¶

Public Access¶

For public APIs (default):

--allow-unauthenticated

Authenticated Access¶

For secured APIs:

--no-allow-unauthenticated

Access with authentication:

# Get identity token
TOKEN=$(gcloud auth print-identity-token)

# Make authenticated request
curl -H "Authorization: Bearer $TOKEN" https://YOUR-SERVICE-URL/health

CI/CD Integration¶

GitHub Actions¶

name: Deploy to Cloud Run

on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Authenticate to GCP
        uses: google-github-actions/auth@v2
        with:
          credentials_json: ${{ secrets.GCP_SA_KEY }}

      - name: Set up Cloud SDK
        uses: google-github-actions/setup-gcloud@v2

      - name: Deploy to Cloud Run
        run: |
          gcloud run deploy schopenhauer-api \
            --source . \
            --region us-central1 \
            --allow-unauthenticated

Monitoring¶

View Logs¶

gcloud run services logs read schopenhauer-api --region us-central1

View Metrics¶

Visit the Cloud Run console:

https://console.cloud.google.com/run?project=YOUR_PROJECT_ID

Cost Optimization¶

Scale to Zero: Use --min-instances 0 for cost savings
Right-size: Start with 512Mi memory, scale up if needed
Concurrency: Higher concurrency = fewer instances
Cold Start: First request after idle period is slower

Troubleshooting¶

Container fails to start¶

Check logs:

gcloud run services logs read schopenhauer-api --limit 50

Out of memory¶

Increase memory allocation:

gcloud run services update schopenhauer-api --memory 1Gi

Timeout errors¶

Increase timeout:

gcloud run services update schopenhauer-api --timeout 600s