Overview of the Docker Compose Setup

This Docker Compose file defines a comprehensive observability stack, with Promtail playing a crucial role in log aggregation. It orchestrates several services, including Traefik for reverse proxying, Fluent-bit for log forwarding, Alertmanager for alerting, Prometheus for metrics, and Grafana for visualization. The primary goal is to efficiently collect logs from various containers and forward them to Loki for storage and analysis.

Promtail Service Configuration

The promtail service is configured to collect logs and send them to the Loki instance. It uses a custom configuration file (promtail_config.yml) to define how logs should be scraped and labeled. The service exposes ports for its own metrics and for receiving logs via syslog.

Loki Log Aggregation

Loki is the central log aggregation system in this setup. It receives logs from Promtail and makes them available for querying and analysis. The configuration includes persistent storage for logs and rule definitions for processing log streams.

Traefik as a Reverse Proxy

Traefik acts as the entry point for external traffic, routing requests to the appropriate services based on hostnames. It's configured to watch Docker events, automatically discovering and configuring routes for new services. This simplifies access to services like Grafana, Prometheus, and Alertmanager.

Monitoring with Prometheus and Alertmanager

Prometheus is set up to scrape metrics from various services, including Traefik and itself. Alertmanager is configured to receive alerts from Prometheus and manage their routing and deduplication. This provides a robust monitoring and alerting solution.

Data Visualization with Grafana

Grafana is used to visualize both logs from Loki and metrics from Prometheus. It's pre-configured with a data source for Prometheus and includes plugins for enhanced dashboarding. The setup allows for creating dashboards to monitor the health and performance of the entire stack.

External Resources for Docker Compose and Logging

• Docker Compose Documentation
• Promtail Official Documentation
• Loki Official Documentation
• Traefik Documentation
• Prometheus Configuration Documentation

version: '3.7'

services:
  traefik:
    image: traefik:1.7.14
    container_name: traefik
    restart: unless-stopped
    ports:
      - 80:80
    command: >
      --api
      --docker
      --docker.domain=localdns.xyz
      --docker.watch
      --defaultentrypoints=http
      --entrypoints='Name:http Address::80'
      --logLevel=INFO
      --accessLog
      --accessLog.format='json'
      --metrics
      --metrics.prometheus
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
    networks:
      - public
    logging:
      driver: fluentd
      options:
        fluentd-async-connect: "true"
        fluentd-address: 192.168.0.4:24224
        
  fluent-bit:
    image: grafana/fluent-bit-plugin-loki:latest
    container_name: fluent-bit
    restart: unless-stopped
    environment:
      - LOKI_URL=http://loki:3100/loki/api/v1/push
    volumes:
      - ./docker-example/configs/fluent-bit.conf:/fluent-bit/etc/fluent-bit.conf
    ports:
      - "24224:24224"
      - "24224:24224/udp"
    networks:
      - public
    depends_on:
      - loki
    logging:
      driver: "json-file"
      options:
        max-size: "1m"
      
  alertmanager:
    image: prom/alertmanager
    container_name: alertmanager
    restart: unless-stopped
    command:
      - '--config.file=/etc/alertmanager/config.yml'
      - '--storage.path=/alertmanager'
      - '--web.external-url=http://alertmanager.localdns.xyz'
    volumes:
      - ./docker-example/configs/alertmanager.yml:/etc/alertmanager/config.yml
      - ./data/alertmanager:/alertmanager
    networks:
      - public
    depends_on:
      - loki
      - traefik
    labels:
      - traefik.frontend.rule=Host:alertmanager.homedns.xyz
      - traefik.enable=true
      - traefik.port=9093
      - traefik.tags=public
      - traefik.docker.network=public
      - traefik.redirectorservice.frontend.entryPoints=http
    logging:
      driver: fluentd
      options:
        fluentd-async-connect: "true"
        fluentd-address: 192.168.0.4:24224

  prometheus:
    image: prom/prometheus
    container_name: 'prometheus'
    user: root
    restart: unless-stopped
    volumes:
      - ./docker-example/configs/prometheus.yml:/etc/prometheus/prometheus.yml
      - ./docker-example/configs/host_alert_rules.yml:/etc/prometheus/rules/host_alert_rules.yml
      - ./data/prometheus:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--storage.tsdb.retention=30d'
      - '--web.console.libraries=/etc/prometheus/console_libraries'
      - '--web.console.templates=/etc/prometheus/consoles'
      - '--web.external-url=http://prometheus.localdns.xyz'
    networks:
      - public
    depends_on:
      - traefik
    labels:
      - traefik.frontend.rule=Host:prometheus.localdns.xyz
      - traefik.enable=true
      - traefik.port=9090
      - traefik.tags=public
      - traefik.docker.network=public
      - traefik.redirectorservice.frontend.entryPoints=http
    logging:
      driver: fluentd
      options:
        fluentd-async-connect: "true"
        fluentd-address: localhost:24224

  grafana:
    image: grafana/grafana:7.4.2
    container_name: grafana
    restart: unless-stopped
    user: root
    volumes:
      - ./data/grafana:/var/lib/grafana
      - ./docker-example/configs/datasource.yml:/etc/grafana/provisioning/datasources/datasource.yml
    environment:
      - GF_SERVER_ROOT_URL=http://grafana.localdns.xyz
      - GF_SECURITY_ADMIN_USER=${GRAFANA_USER:-admin}
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD:-admin}
      - GF_USERS_ALLOW_SIGN_UP=false
      - GF_INSTALL_PLUGINS=grafana-piechart-panel,grafana-worldmap-panel,camptocamp-prometheus-alertmanager-datasource
      - PROMETHEUS_ENDPOINT=http://prometheus:9090
    networks:
      - public
    ports:
      - 3000:3000
    depends_on:
      - loki
      - traefik
      - prometheus
    labels:
      - traefik.frontend.rule=Host:grafana.localdns.xyz
      - traefik.enable=true
      - traefik.port=3000
      - traefik.tags=public
      - traefik.docker.network=public
      - traefik.redirectorservice.frontend.entryPoints=http
    logging:
      driver: fluentd
      options:
        fluentd-async-connect: "true"
        fluentd-address: 192.168.0.4:24224

  promtail:
    image: grafana/promtail
    container_name: promtail
    command: -config.file=/etc/promtail/config.yml
    volumes:
      #- /var/log:/var/log
      - ./docker-example/configs/promtail_config.yml:/etc/promtail/config.yml
    ports:
      - 9080:9080
      - 1514:1514
      - 1514:1514/udp
    networks:
      - public
    depends_on:
      - loki

  loki:
    image: grafana/loki:2.1.0
    container_name: loki
    command: -config.file=/mnt/loki-local-config.yaml
    user: root
    restart: unless-stopped
    volumes:
      - ./data/loki:/tmp/loki
      - ./docker-example/configs/loki.yml:/mnt/loki-local-config.yaml
      - ./docker-example/configs/loki-rules.yml:/etc/loki/rules/fake/loki-rule.yaml
    ports:
      - 3100:3100
    labels:
      - traefik.frontend.rule=Host:loki.localdns.xyz
      - traefik.enable=true
      - traefik.port=3100
      - traefik.tags=public
      - traefik.docker.network=public
      - traefik.redirectorservice.frontend.entryPoints=http
    networks:
      - public
    depends_on:
      - traefik
    logging:
      driver: "json-file"
      options:
        max-size: "1m"

networks:
  public:
    name: public