docker-compose.yml

services:
  # PostgreSQL database
  # PRODUCTION: Use strong password via POSTGRES_PASSWORD env var instead of "postgres"
  # PRODUCTION: Use managed database service (e.g., Railway, Supabase) instead of self-hosted
  postgres:
    image: pgvector/pgvector:pg15
    container_name: rag_transcript_postgres
    environment:
      POSTGRES_USER: ${POSTGRES_USER:-postgres}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-postgres}
      POSTGRES_DB: ${POSTGRES_DB:-rag_transcript}
    # SECURITY: Ports commented out to prevent external access - only accessible via Docker internal network
    # Uncomment for local development if you need direct database access:
    # ports:
    #   - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U postgres"]
      interval: 10s
      timeout: 5s
      retries: 5

  # Redis (Celery broker/backend)
  # PRODUCTION: Set strong REDIS_PASSWORD env var
  redis:
    image: redis:7.4.7-alpine
    container_name: rag_transcript_redis
    command: redis-server --requirepass ${REDIS_PASSWORD:-changeme}
    # SECURITY: Ports commented out to prevent external access - only accessible via Docker internal network
    # Uncomment for local development if you need direct Redis access:
    # ports:
    #   - "6379:6379"
    volumes:
      - redis_data:/data
    healthcheck:
      test: ["CMD", "redis-cli", "--no-auth-warning", "-a", "${REDIS_PASSWORD:-changeme}", "ping"]
      interval: 10s
      timeout: 5s
      retries: 5

  # Qdrant vector database
  # PRODUCTION: Set strong QDRANT_API_KEY env var
  qdrant:
    image: qdrant/qdrant:v1.16.3
    container_name: rag_transcript_qdrant
    environment:
      - QDRANT__SERVICE__API_KEY=${QDRANT_API_KEY:-changeme}
    ports:
      - "6333:6333"  # HTTP API
      - "6334:6334"  # gRPC
    volumes:
      - qdrant_data:/qdrant/storage
    # Healthcheck disabled - Qdrant doesn't include curl/wget in container
    # healthcheck:
    #   test: ["CMD-SHELL", "wget --no-verbose --tries=1 --spider http://localhost:6333/healthz || exit 1"]
    #   interval: 10s
    #   timeout: 5s
    #   retries: 5

  # FastAPI application
  app:
    build:
      context: ./backend
      dockerfile: Dockerfile
    container_name: rag_transcript_app
    command: uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
    env_file:
      - ./backend/.env
    volumes:
      - ./backend:/app
      - ./storage:/app/storage  # Local storage volume
      - ./hf_cache:/root/.cache/huggingface  # Pre-downloaded HF models
      - ./whisper_cache:/root/.cache/whisper  # Pre-downloaded Whisper models
    ports:
      - "8000:8000"
    environment:
      - DATABASE_URL=postgresql://postgres:postgres@postgres:5432/rag_transcript
      - REDIS_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_BROKER_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_RESULT_BACKEND=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - QDRANT_HOST=qdrant
      - QDRANT_PORT=6333
      - QDRANT_API_KEY=${QDRANT_API_KEY:-changeme}
      - STORAGE_BACKEND=local
      - LOCAL_STORAGE_PATH=/app/storage
      - EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
      - HF_HOME=/root/.cache/huggingface
      - TRANSFORMERS_CACHE=/root/.cache/huggingface
      - SENTENCE_TRANSFORMERS_HOME=/root/.cache/huggingface
      - HF_HUB_OFFLINE=0
      - TRANSFORMERS_OFFLINE=0
      - PYTHONUNBUFFERED=1
      - ENVIRONMENT=${ENVIRONMENT:-development}
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      qdrant:
        condition: service_started

  # Celery worker
  worker:
    build:
      context: ./backend
      dockerfile: Dockerfile
    container_name: rag_transcript_worker
    command: celery -A app.core.celery_app worker --loglevel=info --concurrency=1 --pool=solo
    env_file:
      - ./backend/.env
    volumes:
      - ./backend:/app
      - ./storage:/app/storage
      - ./hf_cache:/root/.cache/huggingface  # Pre-downloaded HF models
      - ./whisper_cache:/root/.cache/whisper  # Pre-downloaded Whisper models
      - ./cookies.txt:/etc/yt-cookies.txt  # YouTube cookies for anti-bot auth (optional)
    environment:
      - DATABASE_URL=postgresql://postgres:postgres@postgres:5432/rag_transcript
      - REDIS_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_BROKER_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_RESULT_BACKEND=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - QDRANT_HOST=qdrant
      - QDRANT_PORT=6333
      - QDRANT_API_KEY=${QDRANT_API_KEY:-changeme}
      - STORAGE_BACKEND=local
      - LOCAL_STORAGE_PATH=/app/storage
      - EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
      - HF_HOME=/root/.cache/huggingface
      - TRANSFORMERS_CACHE=/root/.cache/huggingface
      - SENTENCE_TRANSFORMERS_HOME=/root/.cache/huggingface
      - HF_HUB_OFFLINE=0
      - TRANSFORMERS_OFFLINE=0
      - PYTHONUNBUFFERED=1
      - ENVIRONMENT=${ENVIRONMENT:-development}
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
      qdrant:
        condition: service_started
    restart: on-failure:3

  # Celery beat (scheduled tasks)
  beat:
    build:
      context: ./backend
      dockerfile: Dockerfile
    container_name: rag_transcript_beat
    command: celery -A app.core.celery_app beat --loglevel=info
    env_file:
      - ./backend/.env
    volumes:
      - ./backend:/app
      - ./hf_cache:/root/.cache/huggingface  # Pre-downloaded HF models
      - ./whisper_cache:/root/.cache/whisper  # Pre-downloaded Whisper models
    environment:
      - DATABASE_URL=postgresql://postgres:postgres@postgres:5432/rag_transcript
      - REDIS_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_BROKER_URL=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - CELERY_RESULT_BACKEND=redis://:${REDIS_PASSWORD:-changeme}@redis:6379/0
      - EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
      - HF_HOME=/root/.cache/huggingface
      - TRANSFORMERS_CACHE=/root/.cache/huggingface
      - SENTENCE_TRANSFORMERS_HOME=/root/.cache/huggingface
      - HF_HUB_OFFLINE=0
      - TRANSFORMERS_OFFLINE=0
      - PYTHONUNBUFFERED=1
      - ENVIRONMENT=${ENVIRONMENT:-development}
    depends_on:
      - redis
    restart: on-failure:3
    healthcheck:
      test: ["CMD-SHELL", "celery -A app.core.celery_app inspect ping 2>/dev/null | grep -q 'OK' || ps aux | grep -q '[c]elery.*beat'"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s

  # Next.js frontend
  frontend:
    build:
      context: ./frontend
      dockerfile: Dockerfile
    container_name: rag_transcript_frontend
    command: npm run dev
    env_file:
      - ./frontend/.env
    volumes:
      - ./frontend:/app
      - /app/node_modules  # Preserve container's node_modules
      - /app/.next  # Preserve build cache
    ports:
      - "3000:3000"
    environment:
      - NEXT_PUBLIC_API_URL=http://localhost:8000
      - NEXT_PUBLIC_API_V1_PREFIX=/api/v1
      - BACKEND_INTERNAL_URL=http://app:8000
      - WATCHPACK_POLLING=true  # Enable hot reload in Docker
    depends_on:
      - app

volumes:
  postgres_data:
  redis_data:
  qdrant_data: