README-docker-compose.md

Fashion Analyzer - Docker Compose Setup

This project provides a secure, containerized fashion analysis application using Ollama and FastAPI with Docker Compose.

🏗️ Architecture

The application consists of three services:

1. Ollama Service: Runs the Ollama server with LLaVA model for vision analysis
2. FastAPI Service: Provides the web API and user interface
3. Model Loader: One-time service to download the required LLaVA model

🔒 Security Features

• ✅ Pinned Ollama version (0.9.2) - No critical vulnerabilities
• ✅ Non-root user execution - Enhanced container security
• ✅ Security updates - Latest package updates applied
• ✅ Health checks - Service monitoring and restart policies
• ✅ Network isolation - Services communicate via internal network

🚀 Quick Start

Prerequisites

• Docker Engine 20.10+
• Docker Compose 2.0+

1. Start the Application

# Start all services
docker-compose up -d

# View logs
docker-compose logs -f

# Check service status
docker-compose ps

2. Access the Application

• Web Interface: http://localhost:7860
• API Documentation: http://localhost:7860/docs
• Health Check: http://localhost:7860/health
• Ollama API: http://localhost:11434

3. Stop the Application

# Stop all services
docker-compose down

# Stop and remove volumes (removes downloaded models)
docker-compose down -v

📁 Project Structure

AI/
├── docker-compose.yml      # Main orchestration file
├── Dockerfile.fastapi      # FastAPI service Dockerfile
├── .env                    # Environment variables
├── .dockerignore          # Docker build exclusions
├── fast.py                # FastAPI application
├── requirements.txt       # Python dependencies
└── logs/                  # Application logs (created at runtime)

⚙️ Configuration

Environment Variables (.env)

OLLAMA_HOST=0.0.0.0:11434
OLLAMA_ORIGINS=*
OLLAMA_BASE_URL=http://ollama:11434
OLLAMA_PORT=11434
FASTAPI_PORT=7860

Custom Configuration

To modify ports or other settings:

1. Edit .env file
2. Restart services: docker-compose up -d

🔧 Development

Building Images

# Build only FastAPI service
docker-compose build fastapi

# Build with no cache
docker-compose build --no-cache

Viewing Logs

# All services
docker-compose logs -f

# Specific service
docker-compose logs -f fastapi
docker-compose logs -f ollama

Debugging

# Execute commands in running containers
docker-compose exec fastapi bash
docker-compose exec ollama bash

# Check service health
docker-compose exec fastapi curl http://localhost:7860/health

📊 Monitoring

Health Checks

All services include health checks:

• Ollama: Checks API availability
• FastAPI: Checks application health and Ollama connectivity

Service Dependencies

• FastAPI waits for Ollama to be healthy before starting
• Model loader runs after Ollama is ready

🛠️ Troubleshooting

Common Issues

1. Port conflicts: Change ports in .env file
2. Model download fails: Check internet connection and Ollama logs
3. FastAPI can't connect to Ollama: Verify network configuration

Reset Everything

# Stop and remove everything
docker-compose down -v --remove-orphans

# Remove images
docker-compose down --rmi all

# Start fresh
docker-compose up -d

📈 Scaling

To run multiple FastAPI instances:

# Scale FastAPI service
docker-compose up -d --scale fastapi=3

Note: You'll need a load balancer for multiple instances.

🔐 Security Considerations

• Services run as non-root users
• Network isolation between services
• No sensitive data in environment variables
• Regular security updates applied
• Pinned dependency versions

📝 API Usage

Upload and Analyze Image

curl -X POST "http://localhost:7860/analyze-image" \
     -H "accept: application/json" \
     -H "Content-Type: multipart/form-data" \
     -F "[email protected]"

Health Check

curl http://localhost:7860/health

🤝 Contributing

1. Make changes to the code
2. Test with docker-compose up --build
3. Submit pull request

📄 License

This project is licensed under the MIT License.