monitoring/docker-compose.yml

networks:
  monitoring:
    driver: bridge
  traefik:
    external: true

services:
  grafana:
    image: grafana/grafana:11.2.0
    container_name: grafana
    restart: unless-stopped
    volumes:
      - ./data/grafana:/var/lib/grafana
      - ./conf/grafana/provisioning:/etc/grafana/provisioning
    ports:
      - 3000:3000
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD}
      - GF_USERS_ALLOW_SIGN_UP=false
      - GF_SERVER_ROOT_URL=https://grafana.${PUBLIC_HOST}
      - GF_AUTH_GENERIC_OAUTH_ENABLED=true
      - GF_AUTH_GENERIC_OAUTH_NAME=SSO
      - GF_AUTH_GENERIC_OAUTH_ALLOW_SIGN_UP=true
      - GF_AUTH_GENERIC_OAUTH_CLIENT_ID=grafana
      - GF_AUTH_GENERIC_OAUTH_CLIENT_SECRET=${GRAFANA_CLIENT_SECRET}
      - GF_AUTH_GENERIC_OAUTH_SCOPES=openid email profile offline_access roles
      - GF_AUTH_GENERIC_OAUTH_EMAIL_ATTRIBUTE_PATH=email
      - GF_AUTH_GENERIC_OAUTH_LOGIN_ATTRIBUTE_PATH=preferred_username
      - GF_AUTH_GENERIC_OAUTH_NAME_ATTRIBUTE_PATH=full_name
      - GF_AUTH_GENERIC_OAUTH_AUTH_URL=${OPENID_URL}/protocol/openid-connect/auth
      - GF_AUTH_GENERIC_OAUTH_TOKEN_URL=${OPENID_URL}/protocol/openid-connect/token
      - GF_AUTH_GENERIC_OAUTH_API_URL=${OPENID_URL}/protocol/openid-connect/userinfo
      - GF_AUTH_GENERIC_OAUTH_ROLE_ATTRIBUTE_PATH=contains(roles[*], 'admin') && 'Admin' || contains(roles[*], 'editor') && 'Editor' || 'Viewer'
    networks:
      - monitoring
      - traefik
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.grafana.rule=Host(`grafana.${HOST_SUFFIX}`)"
      - "traefik.http.routers.grafana.entrypoints=web"
  prometheus:
    image: prom/prometheus:v2.54.1
    container_name: prometheus
    restart: unless-stopped
    volumes:
      - ./conf/prometheus:/etc/prometheus
      - ./data/prometheus:/prometheus
    ports:
      - 9090:9090
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--storage.tsdb.retention.time=120d'
      - '--web.console.libraries=/etc/prometheus/console_libraries'
      - '--web.console.templates=/etc/prometheus/consoles'
      - '--web.enable-lifecycle'
      - "--web.external-url=http://prometheus.${HOST_SUFFIX}"
    expose:
      - 9090
    networks:
      - monitoring
      - traefik
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.prometheus.rule=Host(`prometheus.${HOST_SUFFIX}`)"
      - "traefik.http.routers.prometheus.entrypoints=web"
  alertmanager:
    image: prom/alertmanager:v0.27.0
    container_name: alertmanager
    restart: unless-stopped
    volumes:
      - ./conf/alertmanager:/etc/alertmanager
      - ./data/alertmanager:/alertmanager
    environment:
      - TZ=${TZ}
    ports:
      - 9093:9093
    command:
      - '--config.file=/etc/alertmanager/alertmanager.yml'
      - '--storage.path=/alertmanager'
      - "--web.external-url=http://alertmanager.${HOST_SUFFIX}"
    expose:
      - 9093
    networks:
      - monitoring
      - traefik
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.alertmanager.rule=Host(`alertmanager.${HOST_SUFFIX}`)"
      - "traefik.http.routers.alertmanager.entrypoints=web"
  node-exporter:
    image: prom/node-exporter:v1.8.2
    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)($$|/)'
    expose:
      - 9100
    networks:
      - monitoring
  blackbox_exporter:
    image: prom/blackbox-exporter:v0.25.0
    container_name: blackbox
    restart: unless-stopped
    expose:
      - 9115
    networks:
      - monitoring
  # smokeping:
  #   image: quay.io/superq/smokeping-prober
  #   container_name: smokeping
  #   restart: unless-stopped
  #   command: nyyu.dev
  #   privileged: true
  #   expose:
  #     - 9374
  #   networks:
  #     - monitoring
  cadvisor:
    # ARM image
    image: gcr.io/cadvisor/cadvisor:v0.50.0
    container_name: cadvisor
    restart: unless-stopped
    command:
      - '--housekeeping_interval=10s'
      - '--raw_cgroup_prefix_whitelist=/docker/'
      - '--disable_metrics=cpu_topology,hugetlb'
    privileged: true
    pid: 'host'
    ports:
      - '8040:8080'
    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:/dev/kmsg'
    expose:
      - 8080
    networks:
      - monitoring