docker-compose.yml

# ============================================================
# My AI Learning Notes - Docker Compose 配置
# ============================================================
#
# 快速啟動所有服務：
#   docker-compose up -d
#
# 啟動特定服務：
#   docker-compose up -d chromadb ollama
#
# 查看日誌：
#   docker-compose logs -f
#
# 停止所有服務：
#   docker-compose down
#
# 停止並刪除數據：
#   docker-compose down -v
#
# ============================================================

version: '3.8'

services:
  # ==================== 向量數據庫 ====================

  # ChromaDB - 輕量級向量數據庫
  chromadb:
    image: chromadb/chroma:latest
    container_name: ai-learning-chromadb
    ports:
      - "8000:8000"
    volumes:
      - chromadb_data:/chroma/chroma
    environment:
      - IS_PERSISTENT=TRUE
      - ANONYMIZED_TELEMETRY=FALSE
    networks:
      - ai-network
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/api/v1/heartbeat"]
      interval: 30s
      timeout: 10s
      retries: 3

  # Qdrant - 高性能向量數據庫
  qdrant:
    image: qdrant/qdrant:latest
    container_name: ai-learning-qdrant
    ports:
      - "6333:6333"
      - "6334:6334"  # gRPC port
    volumes:
      - qdrant_data:/qdrant/storage
    environment:
      - QDRANT__SERVICE__GRPC_PORT=6334
    networks:
      - ai-network
    restart: unless-stopped

  # ==================== 本地 LLM ====================

  # Ollama - 本地運行大型語言模型
  ollama:
    image: ollama/ollama:latest
    container_name: ai-learning-ollama
    ports:
      - "11434:11434"
    volumes:
      - ollama_data:/root/.ollama
    environment:
      - OLLAMA_HOST=0.0.0.0
    networks:
      - ai-network
    restart: unless-stopped
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    # 如果沒有 GPU，註解掉上面的 deploy 部分

  # Ollama Web UI（可選）
  ollama-webui:
    image: ghcr.io/open-webui/open-webui:main
    container_name: ai-learning-ollama-webui
    ports:
      - "3000:8080"
    volumes:
      - ollama_webui_data:/app/backend/data
    environment:
      - OLLAMA_BASE_URL=http://ollama:11434
    depends_on:
      - ollama
    networks:
      - ai-network
    restart: unless-stopped

  # ==================== 數據庫 ====================

  # PostgreSQL - 關係型數據庫（支援 pgvector）
  postgres:
    image: pgvector/pgvector:pg16
    container_name: ai-learning-postgres
    ports:
      - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
    environment:
      - POSTGRES_DB=ai_learning
      - POSTGRES_USER=ai_user
      - POSTGRES_PASSWORD=ai_password_change_in_production
    networks:
      - ai-network
    restart: unless-stopped
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ai_user -d ai_learning"]
      interval: 10s
      timeout: 5s
      retries: 5

  # MongoDB - NoSQL 數據庫
  mongodb:
    image: mongo:7
    container_name: ai-learning-mongodb
    ports:
      - "27017:27017"
    volumes:
      - mongodb_data:/data/db
    environment:
      - MONGO_INITDB_ROOT_USERNAME=admin
      - MONGO_INITDB_ROOT_PASSWORD=admin_password_change_in_production
      - MONGO_INITDB_DATABASE=ai_learning
    networks:
      - ai-network
    restart: unless-stopped

  # Redis - 快取和消息隊列
  redis:
    image: redis:7-alpine
    container_name: ai-learning-redis
    ports:
      - "6379:6379"
    volumes:
      - redis_data:/data
    command: redis-server --appendonly yes --requirepass redis_password_change_in_production
    networks:
      - ai-network
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "redis-cli", "--raw", "incr", "ping"]
      interval: 10s
      timeout: 3s
      retries: 5

  # ==================== 搜尋引擎 ====================

  # Elasticsearch - 全文搜尋引擎
  elasticsearch:
    image: docker.elastic.co/elasticsearch/elasticsearch:8.11.0
    container_name: ai-learning-elasticsearch
    ports:
      - "9200:9200"
      - "9300:9300"
    volumes:
      - elasticsearch_data:/usr/share/elasticsearch/data
    environment:
      - discovery.type=single-node
      - xpack.security.enabled=false
      - "ES_JAVA_OPTS=-Xms512m -Xmx512m"
    networks:
      - ai-network
    restart: unless-stopped

  # ==================== 監控與追蹤 ====================

  # MLflow - 機器學習實驗追蹤
  mlflow:
    image: ghcr.io/mlflow/mlflow:latest
    container_name: ai-learning-mlflow
    ports:
      - "5000:5000"
    volumes:
      - mlflow_data:/mlflow
    command: >
      mlflow server
      --host 0.0.0.0
      --port 5000
      --backend-store-uri sqlite:///mlflow/mlflow.db
      --default-artifact-root /mlflow/artifacts
    networks:
      - ai-network
    restart: unless-stopped

  # Prometheus - 指標收集
  prometheus:
    image: prom/prometheus:latest
    container_name: ai-learning-prometheus
    ports:
      - "9090:9090"
    volumes:
      - ./monitoring/prometheus.yml:/etc/prometheus/prometheus.yml
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
    networks:
      - ai-network
    restart: unless-stopped

  # Grafana - 數據可視化
  grafana:
    image: grafana/grafana:latest
    container_name: ai-learning-grafana
    ports:
      - "3001:3000"
    volumes:
      - grafana_data:/var/lib/grafana
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=admin_change_in_production
      - GF_USERS_ALLOW_SIGN_UP=false
    depends_on:
      - prometheus
    networks:
      - ai-network
    restart: unless-stopped

  # ==================== 消息隊列 ====================

  # RabbitMQ - 消息代理
  rabbitmq:
    image: rabbitmq:3-management-alpine
    container_name: ai-learning-rabbitmq
    ports:
      - "5672:5672"   # AMQP port
      - "15672:15672" # Management UI
    volumes:
      - rabbitmq_data:/var/lib/rabbitmq
    environment:
      - RABBITMQ_DEFAULT_USER=admin
      - RABBITMQ_DEFAULT_PASS=admin_password_change_in_production
    networks:
      - ai-network
    restart: unless-stopped

  # ==================== Web 應用範例 ====================

  # Jupyter Lab - 開發環境
  jupyter:
    build:
      context: .
      dockerfile: docker/Dockerfile.jupyter
    container_name: ai-learning-jupyter
    ports:
      - "8888:8888"
    volumes:
      - ./:/workspace
      - jupyter_data:/root/.jupyter
    environment:
      - JUPYTER_ENABLE_LAB=yes
      - JUPYTER_TOKEN=your-token-change-in-production
    networks:
      - ai-network
    restart: unless-stopped
    # 如果需要 GPU
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - driver: nvidia
    #           count: all
    #           capabilities: [gpu]

# ==================== 網絡配置 ====================
networks:
  ai-network:
    driver: bridge
    ipam:
      config:
        - subnet: 172.28.0.0/16

# ==================== 數據卷配置 ====================
volumes:
  # 向量數據庫
  chromadb_data:
  qdrant_data:

  # LLM
  ollama_data:
  ollama_webui_data:

  # 數據庫
  postgres_data:
  mongodb_data:
  redis_data:
  elasticsearch_data:

  # 監控
  mlflow_data:
  prometheus_data:
  grafana_data:

  # 消息隊列
  rabbitmq_data:

  # 開發環境
  jupyter_data: