docker-compose.yml

# Docker Compose v3 — совместим и с `docker compose`, (частично) и с `docker stack deploy`.
# Для swarm использовать нужно docker-stack.yml (учитывает global-режим exporters и constraints).
#
# Перед запуском убедиться, что есть внешняя (overlay) сеть в кластере:
#   docker network create --driver overlay --attachable tool-net
#
# У каждого сервиса своя папка конфигов в этом каталоге.

version: "3.8"

networks:
  tool-net:
    external: true

volumes:
  minio-data:
  postgres-data:
  grafana-data:
  prometheus-data:
  portainer_data:
  redis-data:
  caddy_data:
  caddy_config:

services:
  # ------------------------------------------------------------------
  # Caddy — ingress / reverse proxy (единственный публикуется наружу)
  # ------------------------------------------------------------------
  caddy:
    image: caddy:latest
    container_name: caddy
    networks:
      - tool-net
    ports:
      - "80:80"
      - "443:443"
    volumes:
      - ./caddy/Caddyfile:/etc/caddy/Caddyfile:ro
      - caddy_data:/data
      - caddy_config:/config
    restart: unless-stopped

  # ------------------------------------------------------------------
  # Grafana — дашборды и визуализация
  # ------------------------------------------------------------------
  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    networks:
      - tool-net
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=admin
      - GF_USERS_ALLOW_SIGN_UP=false
    volumes:
      - ./grafana/provisioning:/etc/grafana/provisioning:ro
      - ./grafana/dashboards:/var/lib/grafana/dashboards:ro
      - grafana-data:/var/lib/grafana
    restart: unless-stopped

  # ------------------------------------------------------------------
  # Prometheus — сбор и хранение метрик
  # ------------------------------------------------------------------
  prometheus:
    image: prom/prometheus:latest
    container_name: prometheus
    networks:
      - tool-net
    command:
      - --config.file=/etc/prometheus/prometheus.yml
      - --storage.tsdb.path=/prometheus
    volumes:
      - ./prometheus/prometheus.yml:/etc/prometheus/prometheus.yml:ro
      - prometheus-data:/prometheus
    restart: unless-stopped

  # ------------------------------------------------------------------
  # Node Exporter — метрики хоста (в swarm: global mode, см. stack)
  # ------------------------------------------------------------------
  node-exporter:
    image: prom/node-exporter:latest
    container_name: node-exporter
    networks:
      - tool-net
    command:
      - --path.rootfs=/host
    pid: host
    volumes:
      - /:/host:ro,rslave
    restart: unless-stopped

  # ------------------------------------------------------------------
  # cAdvisor — метрики контейнеров (в swarm: global mode, см. stack)
  # ------------------------------------------------------------------
  cadvisor:
    image: gcr.io/cadvisor/cadvisor:latest
    container_name: cadvisor
    networks:
      - tool-net
    volumes:
      - /:/rootfs:ro
      - /var/run:/var/run:ro
      - /sys:/sys:ro
      - /var/lib/docker/:/var/lib/docker:ro
      - /dev/disk/:/dev/disk:ro
    devices:
      - /dev/kmsg
    restart: unless-stopped

  # ------------------------------------------------------------------
  # MLflow — tracking server (бэкенд: PostgreSQL, артефакты: MinIO)
  # ------------------------------------------------------------------
  mlflow:
    image: ghcr.io/mlflow/mlflow:latest
    container_name: mlflow
    networks:
      - tool-net
    command:
      - mlflow
      - server
      - --host=0.0.0.0
      - --port=5000
      - --backend-store-uri=postgresql://mlflow:mlflow@postgres:5432/mlflow
      - --default-artifact-root=s3://mlflow
    env_file:
      - ./mlflow/mlflow.env
    depends_on:
      - postgres
      - minio
    restart: unless-stopped

  # ------------------------------------------------------------------
  # MinIO — S3-совместимое хранилище артефактов
  # ------------------------------------------------------------------
  minio:
    image: minio/minio:latest
    container_name: minio
    networks:
      - tool-net
    command: server /data --console-address ":9001"
    env_file:
      - ./minio/minio.env
    volumes:
      - minio-data:/data
    healthcheck:
      test: ["CMD", "mc", "ready", "local"]
      interval: 5s
      timeout: 5s
      retries: 5
    restart: unless-stopped

  # Init-контейнер: создаёт бакет mlflow в MinIO
  minio-init:
    image: minio/mc:latest
    container_name: minio-init
    networks:
      - tool-net
    entrypoint: ["/bin/sh", "-c"]
    command:
      - >
        until mc alias set local http://minio:9000 minioadmin minioadmin; do sleep 2; done;
        mc mb --ignore-existing local/mlflow;
    env_file:
      - ./minio/minio.env
    depends_on:
      - minio
    restart: "no"

  # ------------------------------------------------------------------
  # Portainer — управление Docker/Swarm
  # ------------------------------------------------------------------
  portainer:
    image: portainer/portainer-ce:latest
    container_name: portainer
    networks:
      - tool-net
    command: -H unix:///var/run/docker.sock
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - portainer-data:/data
    restart: unless-stopped

  # ------------------------------------------------------------------
  # Adminer — управление БД (postgres)
  # ------------------------------------------------------------------
  adminer:
    image: adminer:latest
    container_name: adminer
    networks:
      - tool-net
    environment:
      - ADMINER_DEFAULT_SERVER=postgres
      - ADMINER_DEFAULT_USER=mlflow
      - ADMINER_DEFAULT_PASSWORD=mlflow
      - ADMINER_DEFAULT_DB=mlflow
    depends_on:
      - postgres
    restart: unless-stopped

  # ------------------------------------------------------------------
  # PostgreSQL — БД для MLflow (и общая на будущее)
  # ------------------------------------------------------------------
  postgres:
    image: postgres:latest
    container_name: postgres
    networks:
      - tool-net
    environment:
      - POSTGRES_USER=postgres
      - POSTGRES_PASSWORD=postgres
      - POSTGRES_DB=postgres
    volumes:
      - ./postgres/initdb:/docker-entrypoint-initdb.d:ro
      - postgres-data:/var/lib/postgresql/data
    restart: unless-stopped

  # ------------------------------------------------------------------
  # Redis — закомментирован (активируйте при необходимости)
  # ------------------------------------------------------------------
  # redis:
  #   image: redis:latest
  #   container_name: redis
  #   networks:
  #     - tool-net
  #   command: redis-server --appendonly yes
  #   volumes:
  #     - redis-data:/data
  #   restart: unless-stopped

  # ------------------------------------------------------------------
  # Prometheus Alertmanager — уведомления (закомментирован, опционально)
  # ------------------------------------------------------------------
  # alertmanager:
  #   image: prom/alertmanager:latest
  #   container_name: alertmanager
  #   networks:
  #     - tool-net
  #   volumes:
  #     - ./prometheus/alertmanager.yml:/etc/alertmanager/alertmanager.yml:ro
  #   restart: unless-stopped