docker-compose.yml

version: '3.8'

services:
  # PostgreSQL Database
  postgres:
    image: postgres:15-alpine
    container_name: astroml-postgres
    environment:
      POSTGRES_DB: astroml
      POSTGRES_USER: astroml
      POSTGRES_PASSWORD: astroml_password
      POSTGRES_INITDB_ARGS: "--encoding=UTF-8 --lc-collate=C --lc-ctype=C"
    ports:
      - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
      - ./migrations:/docker-entrypoint-initdb.d
    networks:
      - astroml-network
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U astroml -d astroml"]
      interval: 10s
      timeout: 5s
      retries: 5
    restart: unless-stopped

  # Redis for caching and job queues
  redis:
    image: redis:7-alpine
    container_name: astroml-redis
    ports:
      - "6379:6379"
    volumes:
      - redis_data:/data
    networks:
      - astroml-network
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 5s
      retries: 5
    restart: unless-stopped
    command: redis-server --appendonly yes

  # Feature Store Service
  feature-store:
    build:
      context: .
      target: feature-store
    container_name: astroml-feature-store
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - LOG_LEVEL=INFO
      - ASTROML_ENV=container
    ports:
      - "8000:8000"
      - "8080:8080"
    volumes:
      - feature_store_data:/app/feature_store
      - feature_store_logs:/app/logs
      - ./config:/app/config:ro
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    restart: unless-stopped
    profiles:
      - feature-store
      - full

  # Ingestion Service
  ingestion:
    build:
      context: .
      target: ingestion
    container_name: astroml-ingestion
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - LOG_LEVEL=INFO
      - STELLAR_NETWORK_PASSPHRASE=Public Global Stellar Network ; September 2015
      - ASTROML_ENV=container
    ports:
      - "8001:8000"
      - "8081:8080"
    volumes:
      - ./config:/app/config:ro
      - ingestion_logs:/app/logs
      - ingestion_data:/app/data
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: unless-stopped
    command: ["python", "-m", "astroml.ingestion"]

  # Enhanced Streaming Service
  streaming:
    build:
      context: .
      target: ingestion
    container_name: astroml-streaming
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - LOG_LEVEL=INFO
      - STELLAR_HORIZON_URL=https://horizon.stellar.org
      - STELLAR_NETWORK_PASSPHRASE=Public Global Stellar Network ; September 2015
      - ASTROML_ENV=container
    ports:
      - "8002:8000"
    volumes:
      - ./config:/app/config:ro
      - streaming_logs:/app/logs
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: unless-stopped
    command: ["python", "-m", "astroml.ingestion.enhanced_stream"]

  # Training Service (GPU-enabled)
  training-gpu:
    build:
      context: .
      target: training
    container_name: astroml-training-gpu
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - CUDA_VISIBLE_DEVICES=0
      - PYTHONPATH=/app
      - ASTROML_ENV=container
    ports:
      - "6006:6006"  # TensorBoard
    volumes:
      - ./config:/app/config:ro
      - training_models:/app/models
      - training_data:/app/data
      - training_logs:/app/logs
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: "no"  # Training jobs are typically run once
    profiles:
      - gpu
      - full
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]

  # Training Service (CPU-only)
  training-cpu:
    build:
      context: .
      target: training-cpu
    container_name: astroml-training-cpu
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - PYTHONPATH=/app
      - ASTROML_ENV=container
    ports:
      - "6007:6006"  # TensorBoard
    volumes:
      - ./config:/app/config:ro
      - training_models:/app/models
      - training_data:/app/data
      - training_logs:/app/logs
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: "no"
    profiles:
      - cpu
      - full

  # Development Environment
  dev:
    build:
      context: .
      target: development
    container_name: astroml-dev
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - PYTHONPATH=/app
      - ASTROML_ENV=container
    ports:
      - "8003:8000"
      - "8888:8888"  # Jupyter
      - "6008:6006"  # TensorBoard
    volumes:
      - .:/app
      - dev_logs:/app/logs
      - dev_data:/app/data
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: unless-stopped
    profiles:
      - dev
      - full
    command: ["jupyter", "lab", "--ip=0.0.0.0", "--port=8888", "--no-browser", "--allow-root"]

  # Production Service
  production:
    build:
      context: .
      target: production
    container_name: astroml-production
    environment:
      - DATABASE_URL=postgresql://astroml:astroml_password@postgres:5432/astroml
      - REDIS_URL=redis://redis:6379/0
      - FEATURE_STORE_PATH=/app/feature_store
      - LOG_LEVEL=WARNING
      - ASTROML_ENV=container
    ports:
      - "8004:8000"
    volumes:
      - ./config:/app/config:ro
      - production_logs:/app/logs
      - production_data:/app/data
      - feature_store_data:/app/feature_store
    networks:
      - astroml-network
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      feature-store:
        condition: service_started
    restart: unless-stopped
    profiles:
      - prod
      - full

  # Monitoring with Prometheus (optional)
  prometheus:
    image: prom/prometheus:latest
    container_name: astroml-prometheus
    ports:
      - "9090:9090"
    volumes:
      - ./monitoring/prometheus.yml:/etc/prometheus/prometheus.yml:ro
      - prometheus_data:/prometheus
    networks:
      - astroml-network
    profiles:
      - monitoring
    restart: unless-stopped

  # Grafana for visualization (optional)
  grafana:
    image: grafana/grafana:latest
    container_name: astroml-grafana
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=admin
    ports:
      - "3000:3000"
    volumes:
      - grafana_data:/var/lib/grafana
      - ./monitoring/grafana/dashboards:/etc/grafana/provisioning/dashboards:ro
      - ./monitoring/grafana/datasources:/etc/grafana/provisioning/datasources:ro
    networks:
      - astroml-network
    depends_on:
      - prometheus
    profiles:
      - monitoring
    restart: unless-stopped

networks:
  astroml-network:
    driver: bridge

volumes:
  postgres_data:
    driver: local
  redis_data:
    driver: local
  feature_store_data:
    driver: local
  feature_store_logs:
    driver: local
  ingestion_logs:
    driver: local
  ingestion_data:
    driver: local
  streaming_logs:
    driver: local
  training_models:
    driver: local
  training_data:
    driver: local
  training_logs:
    driver: local
  dev_logs:
    driver: local
  dev_data:
    driver: local
  production_logs:
    driver: local
  production_data:
    driver: local
  prometheus_data:
    driver: local
  grafana_data:
    driver: local