Docker Compose Deployment¶

The simplest way to run NextSight AI locally or in development.

Quick Start¶

git clone https://github.com/nextsight-ai/nextsightai.git
cd nextsight
docker-compose up -d

Access at http://localhost:3000

docker-compose.yml¶

services:
  postgres:
    image: postgres:16-alpine
    container_name: nextsight-postgres
    environment:
      POSTGRES_DB: nextsight
    volumes:
      - postgres_data:/var/lib/postgresql/data

  redis:
    image: redis:7-alpine
    container_name: nextsight-redis

  backend:
    build: ./backend
    container_name: nextsight-backend
    ports:
      - "8000:8000"
    environment:
      - DATABASE_URL=postgresql+asyncpg://postgres:postgres@postgres:5432/nextsight
      - K8S_CONFIG_PATH=/home/nextsight/.kube/config
      - GEMINI_API_KEY=${GEMINI_API_KEY}
    volumes:
      - ~/.kube:/home/nextsight/.kube:ro
    depends_on:
      - postgres
      - redis

  frontend:
    build: ./frontend
    container_name: nextsight-frontend
    ports:
      - "3000:80"
    depends_on:
      - backend

Environment Configuration¶

Create a .env file:

# AI Provider
GEMINI_API_KEY=your-api-key

# Optional
DEBUG=true

Common Commands¶

# Start services
docker-compose up -d

# View logs
docker-compose logs -f

# Stop services
docker-compose down

# Rebuild after changes
docker-compose up -d --build

# Remove volumes
docker-compose down -v

Using Makefile¶

make dev     # Start development
make logs    # View logs
make down    # Stop services
make clean   # Remove containers and volumes

Troubleshooting¶

Port Already in Use¶

# Check what's using the port
lsof -i :3000
lsof -i :8000

# Use different ports (edit docker-compose.yml)
# Change ports mapping in frontend service

Kubernetes Connection Issues¶

Ensure your kubeconfig is accessible:

# Test kubectl access
docker-compose exec nextsight-backend kubectl get nodes

Logs Not Streaming¶

Check WebSocket connectivity:

# Test WebSocket
wscat -c ws://localhost:8000/api/v1/ws/health