Agentic MCP Weather System

# 🚀 Production Deployment Guide - Weather Intelligence System ## **🌟 Current System Architecture** Your system includes: - **🌐 Streamlit Chat Interface** (Port 8501) - Primary user interface - **🔧 Weather API + Agent Hub** (Port 8000) - REST API with multi-agent coordination - **🤖 Ollama LLM Engine** (Port 11434) - Local AI models for intelligent responses - **📊 Health Monitoring** - Comprehensive system status and performance tracking --- ## **⚡ Quick Production Setup (Docker - Recommended)** ### Prerequisites - Docker 20.10+ with Docker Compose - 8GB+ RAM (for AI models) - 15GB+ disk space - Ports: 8000, 8501, 11434 ### One-Command Production Deployment ```bash # Clone and deploy git clone <your-repo-url> weather-production cd weather-production # Production deployment with security enabled ./start-docker.sh --prod --build # Verify all services are running docker ps curl http://localhost:8000/health ``` **✅ Production System Ready:** - 🌐 **User Interface**: https://yourdomain.com:8501 (Streamlit Chat) - 🔧 **API Endpoint**: https://yourdomain.com:8000 (REST API + Docs) - 📊 **Health Monitoring**: https://yourdomain.com:8000/health - 🤖 **AI Models**: Ollama with llama3 and phi3 models --- ## **🔧 Advanced Production Configuration** ### Environment Setup ```bash # Create production environment configuration cp .env.example .env.production # Edit production settings (REQUIRED) nano .env.production ``` **Key Production Settings:** ```bash # Server Configuration ENVIRONMENT=production SERVER_HOST=0.0.0.0 SERVER_PORT=8000 # Security Settings (REQUIRED) API_KEY_REQUIRED=true API_KEY=your-secure-production-api-key-here RATE_LIMIT_PER_MINUTE=100 ALLOWED_ORIGINS=https://yourdomain.com # Performance Optimization OLLAMA_MODEL=llama3 MAX_CONCURRENT_REQUESTS=50 REQUEST_TIMEOUT=30.0 # Logging & Monitoring LOG_LEVEL=INFO LOG_FILE_PATH=/var/log/weather-intelligence/server.log ``` ### SSL/HTTPS Configuration ```bash # Add SSL certificates mkdir -p ./ssl # Place your cert.pem and key.pem in ./ssl/ # Update environment for HTTPS echo "SSL_CERT_PATH=/app/ssl/cert.pem" >> .env.production echo "SSL_KEY_PATH=/app/ssl/key.pem" >> .env.production ``` --- ## **🐳 Docker Production Deployment Options** ### Option 1: Complete System (Recommended) ```bash # Full system with Streamlit + API + Ollama docker-compose --env-file .env.production up -d --build # Verify deployment docker-compose ps docker-compose logs --tail=50 ``` ### Option 2: Production with Custom Configuration ```bash # Use production overrides docker-compose -f docker-compose.yml -f docker-compose.prod.yml up -d # Check service health curl https://yourdomain.com:8000/health curl https://yourdomain.com:8501 # Streamlit interface ``` ### Option 3: Scaling for High Traffic ```bash # Scale weather server instances docker-compose up -d --scale weather-server=3 # Load balancer configuration (add nginx/traefik) # See advanced deployment section below ``` docker-compose logs -f weather-server ``` #### Option B: Docker with External Ollama ```bash # Ensure Ollama is running on host ollama serve & # Build production image docker build -t weather-mcp:latest . # Run with environment file (connecting to host Ollama) docker run -d \ --name weather-mcp \ --env-file .env \ -p 8000:8000 \ --add-host=host.docker.internal:host-gateway \ --restart unless-stopped \ weather-mcp:latest # Check container status docker logs weather-mcp ``` ### 3. Direct Python Deployment ```bash # Install dependencies pip install -r requirements.txt # Validate configuration python main.py validate # Start production server python main.py start ``` ### 4. Production with Reverse Proxy (Nginx) ```nginx server { listen 80; server_name your-domain.com; location / { proxy_pass http://localhost:8000; proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header X-Forwarded-Proto $scheme; } location /health { proxy_pass http://localhost:8000/health; access_log off; } } ``` ## Production Checklist ### Before Deployment - [ ] Install and configure Ollama with Llama3 model - [ ] Verify Ollama is running: `ollama serve` - [ ] Test Ollama connectivity: `ollama list` - [ ] Set `ENVIRONMENT=production` in `.env` - [ ] Configure secure `API_KEY` if authentication required - [ ] Set appropriate `ALLOWED_ORIGINS` for CORS - [ ] Configure log file path and rotation - [ ] Set resource limits (`MAX_CONNECTIONS`, `RATE_LIMIT_PER_MINUTE`) - [ ] Test configuration: `python main.py validate` ### Security Hardening - [ ] Enable API key authentication (`API_KEY_REQUIRED=true`) - [ ] Restrict CORS origins to your domain only - [ ] Set up SSL/TLS certificates - [ ] Configure firewall rules (allow only necessary ports) - [ ] Set up log monitoring and alerts - [ ] Regular security updates ### Monitoring Setup ```bash # Health check endpoint for monitoring curl http://localhost:8000/health/quick # Comprehensive health check curl http://localhost:8000/health # Server metrics curl http://localhost:8000/info ``` ### Performance Tuning ```bash # Multiple workers for high traffic uvicorn weather:app --host 0.0.0.0 --port 8000 --workers 4 # With Gunicorn (alternative) gunicorn weather:app -w 4 -k uvicorn.workers.UvicornWorker --bind 0.0.0.0:8000 ``` ## Environment Variables Reference | Variable | Required | Default | Description | |----------|----------|---------|-------------| | `ENVIRONMENT` | No | `production` | Deployment environment | | `SERVER_HOST` | No | `0.0.0.0` | Server bind address | | `SERVER_PORT` | No | `8000` | Server port | | `API_KEY_REQUIRED` | No | `false` | Enable API authentication | | `API_KEY` | If auth enabled | - | API key for authentication | | `LOG_LEVEL` | No | `INFO` | Logging level | | `RATE_LIMIT_PER_MINUTE` | No | `100` | API rate limit | | `OLLAMA_HOST` | No | `http://localhost:11434` | Ollama server URL | | `OLLAMA_MODEL` | No | `llama3` | Ollama model to use | | `OLLAMA_TIMEOUT` | No | `30.0` | Ollama request timeout | | `ENABLE_ADVANCED_ORCHESTRATION` | No | `true` | Enable LLM features | See `.env.example` for complete list. ## Troubleshooting ### Common Issues 1. **Port Already in Use** ```bash lsof -i :8000 # Kill the process or use a different port ``` 2. **Permission Denied (Docker)** ```bash # Run as non-root user docker run --user $(id -u):$(id -g) ... ``` 3. **External API Timeouts** ```bash # Increase timeout in .env SERVER_TIMEOUT=45.0 ``` 4. **Ollama Not Responding** ```bash # Check if Ollama is running ps aux | grep ollama # Restart Ollama ollama serve # Test Ollama API curl http://localhost:11434/api/version # Check available models ollama list ``` ### Log Analysis ```bash # Real-time log monitoring tail -f /var/log/weather-mcp/server.log # Error analysis grep -i "error\|exception" /var/log/weather-mcp/server.log # Performance monitoring grep "response_time" /var/log/weather-mcp/server.log ``` ### Health Monitoring Set up monitoring alerts for: - `/health/quick` endpoint availability - Response time > 5 seconds - Error rate > 1% - Memory/CPU usage > 80% ## Scaling Considerations - **Horizontal scaling**: Deploy multiple instances behind load balancer - **Database**: Add Redis for caching weather data - **CDN**: Cache static responses for common locations - **Rate limiting**: Implement distributed rate limiting for multiple instances