docker-compose.yml

# =============================================================================
# OmniDiag — Docker Compose (v2.0 — Full MLOps Stack)
# =============================================================================
# Services:
#   postgres    — PostgreSQL 15 database (primary datastore)
#   redis       — Redis 7 cache
#   backend     — FastAPI application (port 7860)
#   mlflow      — MLflow tracking server (port 5000)
#   prometheus  — Prometheus metrics scraper (port 9090)
#   grafana     — Grafana dashboard (port 3001)
#   retrain     — One-shot retrain container (runs on demand)
#
# Usage:
#   docker compose up -d postgres redis                     # infra only
#   docker compose up -d                                    # full stack
#   docker compose run --rm retrain --disease heart_disease # manual retrain
#   docker compose down -v                                  # wipe volumes
# =============================================================================

services:

  # ── PostgreSQL Database ──────────────────────────────────────────────────────
  postgres:
    image: postgres:15-alpine
    container_name: omnidiag-postgres
    restart: unless-stopped
    environment:
      POSTGRES_USER: omnidiag
      POSTGRES_PASSWORD: omnidiag_pass
      POSTGRES_DB: omnidiag_db
    ports:
      - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U omnidiag"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 10s

  # ── Redis Cache ──────────────────────────────────────────────────────────────
  redis:
    image: redis:7-alpine
    container_name: omnidiag-redis
    restart: unless-stopped
    ports:
      - "6379:6379"
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 5s
      retries: 5

  # ── FastAPI Backend ──────────────────────────────────────────────────────────
  backend:
    build: .
    container_name: omnidiag-backend
    restart: unless-stopped
    ports:
      - "7860:7860"
    environment:
      - DATABASE_URL=postgresql+asyncpg://omnidiag:omnidiag_pass@postgres:5432/omnidiag_db
      - REDIS_URL=redis://redis:6379
      - MLFLOW_TRACKING_URI=http://mlflow:5000
      - CORS_ALLOWED_ORIGINS=http://localhost:5173,http://localhost:3000
      - JWT_SECRET_KEY=${JWT_SECRET_KEY:-change-me-in-production}
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    volumes:
      - .:/app
      - model_cache:/app/models
    command: uvicorn backend.main:app --host 0.0.0.0 --port 7860

  # ── MLflow Tracking Server ───────────────────────────────────────────────────
  mlflow:
    image: python:3.11-slim
    container_name: omnidiag-mlflow
    restart: unless-stopped
    ports:
      - "5000:5000"
    volumes:
      - mlflow_data:/mlruns
    command: >
      sh -c "pip install mlflow psycopg2-binary -q &&
             mlflow server
               --host 0.0.0.0
               --port 5000
               --backend-store-uri sqlite:///mlruns/mlruns.db
               --default-artifact-root /mlruns/artifacts"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:5000/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s

  # ── Prometheus ───────────────────────────────────────────────────────────────
  prometheus:
    image: prom/prometheus:v2.51.0
    container_name: omnidiag-prometheus
    restart: unless-stopped
    ports:
      - "9090:9090"
    volumes:
      - ./deploy/prometheus.yml:/etc/prometheus/prometheus.yml:ro
      - prometheus_data:/prometheus
    command:
      - --config.file=/etc/prometheus/prometheus.yml
      - --storage.tsdb.path=/prometheus
      - --storage.tsdb.retention.time=30d
      - --web.enable-lifecycle
    depends_on:
      - backend

  # ── Grafana ──────────────────────────────────────────────────────────────────
  grafana:
    image: grafana/grafana:10.4.0
    container_name: omnidiag-grafana
    restart: unless-stopped
    ports:
      - "3001:3000"
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD:-omnidiag_grafana}
      - GF_USERS_ALLOW_SIGN_UP=false
      - GF_DASHBOARDS_DEFAULT_HOME_DASHBOARD_PATH=/etc/grafana/dashboards/omnidiag.json
    volumes:
      - grafana_data:/var/lib/grafana
      - ./deploy/grafana/datasources:/etc/grafana/provisioning/datasources:ro
      - ./deploy/grafana/dashboards:/etc/grafana/provisioning/dashboards:ro
      - ./deploy/grafana/dashboards:/etc/grafana/dashboards:ro
    depends_on:
      - prometheus

  # ── Auto-Retrain (on-demand / cron) ─────────────────────────────────────────
  retrain:
    build: .
    container_name: omnidiag-retrain
    profiles:
      - retrain
    environment:
      - DATABASE_URL=postgresql+asyncpg://omnidiag:omnidiag_pass@postgres:5432/omnidiag_db
      - MLFLOW_TRACKING_URI=http://mlflow:5000
    depends_on:
      postgres:
        condition: service_healthy
      mlflow:
        condition: service_healthy
    volumes:
      - .:/app
      - model_cache:/app/models
    entrypoint: ["python", "scripts/retrain.py"]
    command: ["--disease", "heart_disease"]

# ── Named Volumes ─────────────────────────────────────────────────────────────
volumes:
  postgres_data:
    driver: local
  mlflow_data:
    driver: local
  prometheus_data:
    driver: local
  grafana_data:
    driver: local
  model_cache:
    driver: local