docker-compose.yml

networks:
  monitoring:
    driver: bridge

volumes:
  prometheus_data: {}
  grafana_data: {}
  alertmanager_data: {}
  model_serving_logs: {}
  mlflow_data: {}
  mlflow_artifacts: {}

services:
  mlflow-server:
    image: ghcr.io/mlflow/mlflow:v2.21.3
    container_name: mlflow-server
    environment:
      - MLFLOW_ARTIFACT_ROOT=/mlflow/artifacts
      - MLFLOW_BACKEND_STORE_URI=/mlflow/data/mlflow.db
    volumes:
      - mlflow_data:/mlflow/data
      - mlflow_artifacts:/mlflow/artifacts
    ports:
      - "5000:5000"
    networks:
      - default
      - monitoring
    command: >
      mlflow server
      --host 0.0.0.0
      --port 5000
      --backend-store-uri sqlite:///mlflow/data/mlflow.db
      --default-artifact-root /mlflow/artifacts

  node-exporter:
    image: prom/node-exporter:latest
    container_name: node-exporter
    restart: unless-stopped
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - /:/rootfs:ro
    command:
      - '--path.procfs=/host/proc'
      - '--path.rootfs=/rootfs'
      - '--path.sysfs=/host/sys'
      - '--collector.filesystem.mount-points-exclude=^/(sys|proc|dev|host|etc)($$|/)'
    ports:
      - "9100:9100"
    networks:
      - monitoring

  prometheus:
    image: prom/prometheus:latest
    container_name: prometheus
    restart: unless-stopped
    volumes:
      - ./prometheus/prometheus.yml:/etc/prometheus/prometheus.yml
      - ./prometheus/alert.rules.yml:/etc/prometheus/alert.rules.yml
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--web.console.libraries=/etc/prometheus/console_libraries'
      - '--web.console.templates=/etc/prometheus/consoles'
      - '--web.enable-lifecycle'
    ports:
      - "9090:9090"
    networks:
      - monitoring

  alertmanager:
    image: prom/alertmanager:latest
    container_name: alertmanager
    restart: unless-stopped
    volumes:
      - ./alertmanager/alertmanager.yml:/etc/alertmanager/alertmanager.yml
    ports:
      - 9093:9093
    networks:
      - monitoring

  cadvisor:
    image: gcr.io/cadvisor/cadvisor:latest
    container_name: cadvisor
    restart: unless-stopped
    volumes:
      - /:/rootfs:ro
      - /var/run:/var/run:ro
      - /sys:/sys:ro
      - /var/lib/docker/:/var/lib/docker:ro
      - /dev/disk/:/dev/disk:ro
      - /etc/machine-id:/etc/machine-id:ro
    ports:
      - "8080:8080"
    networks:
      - monitoring

  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    restart: unless-stopped
    ports:
      - "3000:3000"
    volumes:
      - grafana_data:/var/lib/grafana
      - ./grafana/dashboards/system:/etc/grafana/dashboards/system:ro
      - ./grafana/dashboards/api:/etc/grafana/dashboards/api:ro
      - ./grafana/dashboards/ml:/etc/grafana/dashboards/ml:ro
      - ./grafana/provisioning:/etc/grafana/provisioning:ro
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=admin
      - GF_USERS_ALLOW_SIGN_UP=false
      # - GF_INSTALL_PLUGINS=grafana-piechart-panel,grafana-worldmap-panel
    networks:
      - monitoring
    depends_on:
      - prometheus

  train-pipeline:
    build:
      context: .
      dockerfile: Dockerfile-train
    container_name: train-pipeline
    environment:
      - MLFLOW_TRACKING_URI=http://mlflow-server:5000
      - USER=mlflow_user
      - OMP_NUM_THREADS=10
    volumes:
      - .:/app
      - mlflow_artifacts:/mlflow/artifacts
    deploy:
      resources:
        limits:
          cpus: '10'
          memory: 8G
    depends_on:
      - mlflow-server
      - node-exporter
      - prometheus
      - grafana

  model-serving:
    build:
      context: .
      dockerfile: Dockerfile-serve
    container_name: model-serving
    restart: unless-stopped
    environment:
      - MLFLOW_TRACKING_URI=http://mlflow-server:5000
      - MODEL_URI=models:/sk-learn-best-model/latest
      - PYTHONUNBUFFERED=1
      - ENABLE_METRICS=true
    volumes:
      - ./app:/app/app:delegated
      - model_serving_logs:/app/logs
      - mlflow_artifacts:/mlflow/artifacts
      - /dev/log:/dev/log
      - /var/log:/var/log:ro
    deploy:
      resources:
        limits:
          cpus: '10'
          memory: 8G
    command: >
      bash -c "
        mkdir -p /app/logs &&
        ln -sf /dev/stdout /app/logs/stdout.log &&
        ln -sf /dev/stderr /app/logs/stderr.log &&
        uvicorn app.main:app --host 0.0.0.0 --port 5050 --reload --log-config app/logging.conf
      "
    ports:
      - "5050:5050"
    networks:
      - default
      - monitoring
    depends_on:
      mlflow-server:
        condition: service_started
      train-pipeline:
        condition: service_completed_successfully
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:5050/metrics"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 40s
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"

  dcgm-exporter:
    profiles:
      - gpu
    image: nvidia/dcgm-exporter:latest
    container_name: dcgm-exporter
    restart: unless-stopped
    deploy:
      resources:
        limits:
          cpus: '10'
          memory: 2G
    ports:
      - "9400:9400"
    networks:
      - monitoring