docker-compose.yml

# DataPilot Docker Compose
# 
# This file runs the complete DataPilot application:
# - PostgreSQL database
# - FastAPI backend API
# - Chainlit web UI
#
# Usage:
#   docker-compose up -d              # Start all services
#   docker-compose logs -f            # View logs
#   docker-compose down               # Stop all services
#   docker-compose down -v            # Stop and remove volumes

name: datapilot

services:
  # ==========================================================================
  # PostgreSQL Database
  # ==========================================================================
  postgres:
    image: postgres:15-alpine
    container_name: postgres
    logging:
      driver: json-file
      options:
        max-size: "10m"
        max-file: "3"
    environment:
      POSTGRES_USER: datapilot
      POSTGRES_PASSWORD: ${DATABASE_PASSWORD:-datapilot_dev_password}
      POSTGRES_DB: datapilot_db
      POSTGRES_INITDB_ARGS: "--encoding=UTF-8 --no-instructions"
    # Reduce PostgreSQL log verbosity (suppress routine startup/checkpoint details)
    command: ["postgres", "-c", "log_min_messages=error", "-c", "log_checkpoints=off", "-c", "log_connections=off", "-c", "log_disconnections=off"]
    volumes:
      - postgres_data:/var/lib/postgresql/data
    ports:
      - "${POSTGRES_PORT:-5432}:5432"
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U datapilot -d datapilot_db"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 10s
    networks:
      - network
    restart: unless-stopped

  # ==========================================================================
  # FastAPI Backend API
  # ==========================================================================
  backend-server:
    init: true
    build:
      context: .
      dockerfile: docker/Dockerfile
    container_name: backend-server
    logging:
      driver: json-file
      options:
        max-size: "10m"
        max-file: "3"
    environment:
      # Database
      DATABASE_URL: postgresql://datapilot:${DATABASE_PASSWORD:-datapilot_dev_password}@postgres:5432/datapilot_db
      
      # Application
      ENVIRONMENT: ${ENVIRONMENT:-production}
      LOG_LEVEL: ${LOG_LEVEL:-INFO}
      DEBUG: ${DEBUG:-false}
      SECRET_KEY: ${SECRET_KEY:-change-this-in-production}
      
      # Data paths (inside container)
      SNAPDATA_PATH: /data
      
      # LLM Configuration
      LLM_PROVIDER: ${LLM_PROVIDER:-ollama}
      OPENAI_API_KEY: ${OPENAI_API_KEY:-}
      ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-}
      OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
      
      # Azure OpenAI Configuration (OpenAI-compatible endpoint)
      AZURE_OPENAI_ENDPOINT: ${AZURE_OPENAI_ENDPOINT:-}
      AZURE_OPENAI_API_KEY: ${AZURE_OPENAI_API_KEY:-}
      
      # LLM Model Settings
      LLM_SQL_MODEL: ${LLM_SQL_MODEL:-}
      LLM_SQL_MAX_TOKENS: ${LLM_SQL_MAX_TOKENS:-2000}
      LLM_SQL_TEMPERATURE: ${LLM_SQL_TEMPERATURE:-0.1}
      LLM_KB_MODEL: ${LLM_KB_MODEL:-}
      LLM_KB_MAX_TOKENS: ${LLM_KB_MAX_TOKENS:-150}
      LLM_KB_MAX_DATA_SIZE: ${LLM_KB_MAX_DATA_SIZE:-8000}
      LLM_KB_MAX_PROMPT_SIZE: ${LLM_KB_MAX_PROMPT_SIZE:-10000}
      LLM_KB_TEMPERATURE: ${LLM_KB_TEMPERATURE:-}
      
      # Query Result Settings
      MAX_RESULT_COLUMNS: ${MAX_RESULT_COLUMNS:-10}
      
      # Vanna Configuration
      SQL_TRAINING_DATA_PATH: ${SQL_TRAINING_DATA_PATH:-./datasets/data/training}
      SYSTEM_PROMPTS_PATH: ${SYSTEM_PROMPTS_PATH:-./datasets/data/prompts}
      VANNA_CHROMADB_PATH: /app/chromadb

      # Disable ChromaDB telemetry and tqdm progress bars
      ANONYMIZED_TELEMETRY: "False"
      TQDM_DISABLE: "1"

      # Suppress noisy library warnings
      PYTHONWARNINGS: "ignore::DeprecationWarning,ignore:urllib3.*doesn't match a supported version"

      # ONNX Runtime optimization for LXC containers (Proxmox, etc.)
      # CRITICAL: Multiple settings to prevent pthread_setaffinity_np errors
      # See: https://github.com/chroma-core/chroma/issues/1420
      OMP_NUM_THREADS: "4"
      ORT_DISABLE_CPU_EP_AFFINITY: "1"
      ORT_DISABLE_THREAD_AFFINITY: "1"
      OMP_WAIT_POLICY: "PASSIVE"
      OMP_PROC_BIND: "false"
      ORT_LOG_LEVEL: "ERROR"
    volumes:
      - datapilot_data:/data
      - ./docker/logs:/app/logs
      - chromadb_data:/app/chromadb
      - chroma_cache:/root/.cache/chroma  # Persist ONNX embedding model (79MB)
    ports:
      - "${API_PORT:-8000}:8000"
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - network
    restart: unless-stopped
    stop_grace_period: 10s
    command: ["uvicorn", "src.api.main:app", "--host", "0.0.0.0", "--port", "8000", "--timeout-keep-alive", "5", "--limit-max-requests", "10000"]
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 10s

  # ==========================================================================
  # Chainlit Web UI
  # ==========================================================================
  frontend-server:
    init: true
    build:
      context: .
      dockerfile: docker/Dockerfile
    container_name: frontend-server
    logging:
      driver: json-file
      options:
        max-size: "10m"
        max-file: "3"
    environment:
      # API connection (internal Docker network)
      API_BASE_URL: http://backend-server:8000
      # External API URL for browser download links
      # Options: "relative" (behind reverse proxy), "http://localhost:8000" (local dev), 
      #          or custom like "https://api.example.com"
      API_EXTERNAL_URL: ${API_EXTERNAL_URL:-http://localhost:8000}
      
      # Required by src.core.config (loaded by chainlit_app.py)
      DATABASE_URL: postgresql://datapilot:${DATABASE_PASSWORD:-datapilot_dev_password}@postgres:5432/datapilot_db
      SECRET_KEY: ${SECRET_KEY:-change-this-in-production}
      ENVIRONMENT: ${ENVIRONMENT:-production}
      LOG_LEVEL: ${LOG_LEVEL:-INFO}
      SNAPDATA_PATH: /data
      
      # LLM Configuration (needed for AI summary generation in Chainlit process)
      LLM_PROVIDER: ${LLM_PROVIDER:-ollama}
      OPENAI_API_KEY: ${OPENAI_API_KEY:-}
      ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-}
      OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
      AZURE_OPENAI_ENDPOINT: ${AZURE_OPENAI_ENDPOINT:-}
      AZURE_OPENAI_API_KEY: ${AZURE_OPENAI_API_KEY:-}
      LLM_SQL_MODEL: ${LLM_SQL_MODEL:-}
      LLM_KB_MODEL: ${LLM_KB_MODEL:-}

      # Chainlit settings
      CHAINLIT_HOST: 0.0.0.0
      CHAINLIT_PORT: 8001
      CHAINLIT_AUTH_SECRET: ${CHAINLIT_AUTH_SECRET:-datapilot-default-auth-secret-change-in-production}
      
      # Suppress verbose logging
      CHAINLIT_DEBUG: "false"

      # Suppress noisy library warnings
      PYTHONWARNINGS: "ignore::DeprecationWarning,ignore:urllib3.*doesn't match a supported version"

      # ONNX Runtime optimization (same as backend)
      OMP_NUM_THREADS: "4"
      ORT_DISABLE_CPU_EP_AFFINITY: "1"
      ORT_DISABLE_THREAD_AFFINITY: "1"
      OMP_WAIT_POLICY: "PASSIVE"
      OMP_PROC_BIND: "false"
    volumes:
      - ./docker/logs:/app/logs
    ports:
      - "${CHAINLIT_PORT:-8001}:8001"
    depends_on:
      backend-server:
        condition: service_healthy
      data-loader:
        condition: service_completed_successfully
    networks:
      - network
    restart: unless-stopped
    stop_grace_period: 10s
    command: ["chainlit", "run", "chainlit_app.py", "--host", "0.0.0.0", "--port", "8001"]
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8001/"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 15s

  # ==========================================================================
  # Data Initialization (downloads data AND loads into database)
  # ==========================================================================
  data-loader:
    init: true
    build:
      context: .
      dockerfile: docker/Dockerfile
    container_name: data-loader
    environment:
      DATABASE_URL: postgresql://datapilot:${DATABASE_PASSWORD:-datapilot_dev_password}@postgres:5432/datapilot_db
      SNAPDATA_PATH: /data
      SECRET_KEY: ${SECRET_KEY:-init-only}

      # ONNX Runtime optimization
      OMP_NUM_THREADS: "4"
      ORT_DISABLE_CPU_EP_AFFINITY: "1"
      ORT_DISABLE_THREAD_AFFINITY: "1"
      OMP_WAIT_POLICY: "PASSIVE"
      OMP_PROC_BIND: "false"
    volumes:
      - datapilot_data:/data
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - network
    command: python /app/scripts/docker_init_data.py

# =============================================================================
# Volumes
# =============================================================================
volumes:
  postgres_data:
    driver: local
  datapilot_data:
    driver: local
  chromadb_data:
    driver: local
  chroma_cache:
    driver: local

# =============================================================================
# Networks
# =============================================================================
networks:
  network:
    driver: bridge