Skip to content

VERCEL_DEPLOYMENT.md

Latest commit

 

History

History
185 lines (150 loc) · 4.67 KB

VERCEL_DEPLOYMENT.md

File metadata and controls

185 lines (150 loc) · 4.67 KB

Vercel Deployment Guide for AstraVideo AI

Environment Variables Setup

1. Frontend Environment Variables

Add these in your Vercel dashboard under Settings > Environment Variables:

NEXT_PUBLIC_API_URL=https://your-app-name.vercel.app/api

2. Backend Environment Variables

Add these in Vercel dashboard:

REDIS_URL=redis://your-redis-url:6379/0
CELERY_BROKER_URL=redis://your-redis-url:6379/1
CELERY_RESULT_BACKEND=redis://your-redis-url:6379/2
JWT_SECRET=your-super-secret-jwt-key-change-this-in-production
OUTPUTS_DIR=/tmp/outputs
MODELSCOPE_CACHE_DIR=/tmp/cache
HUGGINGFACE_CACHE_DIR=/tmp/cache/.huggingface
RATE_LIMIT_REQUESTS=100
RATE_LIMIT_WINDOW=3600
ALLOWED_ORIGINS=https://your-app-name.vercel.app

Required External Services

1. Redis (for Celery queue and caching)

  • Upstash Redis: Free tier available
  • Redis Cloud: Free tier available
  • Redis Labs: Free tier available

2. Worker Service (Separate Deployment)

Since Vercel doesn't support long-running processes like Celery workers, you'll need:

  • Railway: For Celery worker
  • Render: For Celery worker
  • Heroku: For Celery worker
  • DigitalOcean App Platform: For Celery worker

Deployment Steps

1. Connect GitHub Repository

  1. Go to Vercel Dashboard
  2. Click "Add New Project"
  3. Import your GitHub repository: okeylili/AstraVideo-AI
  4. Select the Next.js framework
  5. Set Root Directory to frontend
  6. Set Build Command to npm run build
  7. Set Output Directory to .next

2. Configure Environment Variables

  1. In Vercel project settings, add all environment variables above
  2. Make sure to use Vercel's built-in Redis or external Redis URL

3. Deploy Backend as Serverless Function

  1. Create separate Vercel project for backend
  2. Set Root Directory to backend
  3. Set Build Command to empty (Python doesn't need build)
  4. Set Runtime to Python 3.9+

4. Deploy Worker Service

Deploy the Celery worker to a service that supports long-running processes:

Railway Example:

# Install Railway CLI
npm install -g @railway/cli

# Login and deploy
railway login
railway new
railway set PYTHON_VERSION=3.9
railway set REDIS_URL=your-redis-url
railway deploy

Alternative: Docker Deployment

For a complete deployment with all services, consider using Docker:

1. Create Docker Compose

version: '3.8'
services:
  frontend:
    build: ./frontend
    ports:
      - "3000:3000"
    environment:
      - NEXT_PUBLIC_API_URL=http://backend:8000
    depends_on:
      - backend

  backend:
    build: ./backend
    ports:
      - "8000:8000"
    environment:
      - REDIS_URL=redis://redis:6379/0
      - CELERY_BROKER_URL=redis://redis:6379/1
      - CELERY_RESULT_BACKEND=redis://redis:6379/2
    depends_on:
      - redis

  worker:
    build: ./worker
    environment:
      - REDIS_URL=redis://redis:6379/0
      - CELERY_BROKER_URL=redis://redis:6379/1
      - CELERY_RESULT_BACKEND=redis://redis:6379/2
    depends_on:
      - redis

  redis:
    image: redis:alpine
    ports:
      - "6379:6379"

2. Deploy to Cloud Services

  • AWS ECS: Full container orchestration
  • Google Cloud Run: Serverless containers
  • Azure Container Instances: Simple container deployment
  • DigitalOcean App Platform: Managed containers

Known Limitations

Vercel Limitations

  • No long-running processes: Celery worker cannot run on Vercel
  • Serverless timeout: Maximum 10-30 seconds for API calls
  • File storage: No persistent file system for video outputs
  • Cold starts: May cause delays for first API calls

Workarounds

  1. Video Storage: Use AWS S3, Cloudinary, or Vercel Blob
  2. Worker Service: Deploy to Railway/Render/Heroku
  3. Database: Use managed Redis service
  4. File Uploads: Use temporary storage with CDN

Production Considerations

1. Security

  • Use HTTPS everywhere
  • Implement proper JWT authentication
  • Add rate limiting
  • Validate all inputs

2. Performance

  • Use CDN for static assets
  • Implement caching strategies
  • Monitor API response times
  • Optimize video processing

3. Scaling

  • Horizontal scaling for backend
  • Multiple worker instances
  • Load balancing
  • Auto-scaling based on demand

4. Monitoring

  • Error tracking (Sentry)
  • Performance monitoring
  • Log aggregation
  • Health checks

Quick Start for Development

For local development, use the .env.example file:

cp .env.example .env
# Edit .env with your local settings
npm run dev

Support

If you encounter issues:

  1. Check Vercel deployment logs
  2. Verify environment variables
  3. Test Redis connection
  4. Monitor worker service logs
  5. Check CORS settings