How to Set Up a Monitoring Stack with Prometheus, Grafana, and Node Exporter Using Docker Compose

Monitoring your infrastructure is crucial for reliability and performance. In this article, we'll walk through a simple and effective monitoring stack using Docker Compose, featuring Prometheus for metrics collection, Grafana for visualization, and Node Exporter for exposing host metrics.

Project Structure

.

├── docker-compose.yml

├── prometheus.yml

└── provisioning/

    └── datasources/

        └── prometheus.yml

Understanding the docker-compose.yml File

Let's break down the main components of the docker-compose.yml file:

services:

  prometheus:

    image: prom/prometheus:v3.3.1

    container_name: prometheus

    ports:

      - "9090:9090"

    volumes:

      - ./prometheus.yml:/etc/prometheus/prometheus.yml

      - prometheus_data:/prometheus

    networks:

      - monitoring



  grafana:

    image: grafana/grafana:12.0.0

    container_name: grafana

    ports:

      - "3000:3000"

    environment:

      - GF_SECURITY_ADMIN_PASSWORD=admin

    depends_on:

      - prometheus

    networks:

      - monitoring

    volumes:

      - ./provisioning/datasources:/etc/grafana/provisioning/datasources

      - grafana_data:/var/lib/grafana



  node_exporter:

    image: prom/node-exporter:v1.9.1

    container_name: node_exporter

    ports:

      - "9100:9100"

    networks:

      - monitoring



networks:

  monitoring:

    driver: bridge



volumes:

  prometheus_data:

    driver: local

  grafana_data:

    driver: local

Service Breakdown

• 
  

  Prometheus

  
  
  
  
  
  • Uses the official Prometheus image.
  
  
  • Exposes port 9090.
  
  
  • Mounts a local prometheus.yml config and a Docker volume for persistent data.
  
  
  • Connected to a custom monitoring network.
  
  
  
  
  
  



• 
  
  

  Grafana

  
  
  
  
  
  • Uses the official Grafana image.
  
  
  • Exposes port 3000.
  
  
  • Sets the admin password via environment variable.
  
  
  • Depends on Prometheus (starts after Prometheus is up).
  
  
  • Mounts provisioning files and a persistent data volume.
  
  
  • Connected to the monitoring network.
  
  
  
  
  
  



• 
  
  

  Node Exporter

  
  
  
  
  
  • Uses the official Node Exporter image.
  
  
  • Exposes port 9100.
  
  
  • Connected to the monitoring network.
  
  
  
  
  
  



• 
  
  

  Networks and Volumes

  
  
  
  
  
  • All services share the monitoring bridge network.
  
  
  • Prometheus and Grafana use named volumes for data persistence.
  
  
  
  
  
  

Step-by-Step Guide

1. Clone the Repository

Github Link: https://github.com/rafi021/monitoring-stack-prometheus-grafana-node-exporter

git clone <your-repo-url>

cd <your-repo-directory>

2. Review and Edit Configuration

• 
  prometheus.yml: Make sure Prometheus is configured to scrape Node Exporter.


• 
  provisioning/datasources/prometheus.yml: Ensures Grafana automatically adds Prometheus as a data source.

The prometheus.yml file is the main configuration file for Prometheus. Here’s what each section does:

global:

  scrape_interval: 5s

• 
  global: Sets global configuration options.


• 
  scrape_interval: 5s: Prometheus will scrape (collect) metrics from all configured targets every 5 seconds.
  

scrape_configs:

  - job_name: 'node'

    static_configs:

      - targets: ['node_exporter:9100']  # This points to the Node Exporter container

• 
  scrape_configs: Defines the list of jobs (sets of targets) Prometheus should scrape.


• 
  job_name: 'node': Names this scrape job "node" (for Node Exporter).


• 
  static_configs: Specifies static targets for this job.


• 
  targets: ['node_exporter:9100']: Prometheus will scrape metrics from the Node Exporter container at port 9100.

Summary:


This config tells Prometheus to collect metrics from the Node Exporter container every 5 seconds.

This file provisioning/datasources/prometheus.yml is a Grafana data source provisioning configuration. It tells Grafana to automatically add Prometheus as a data source when Grafana starts.

• apiVersion: 1
  
  
  Specifies the version of the provisioning config format.



• 
  

  datasources:
  
  
  A list of data sources to add.

  
  
  
  
  
  • 
    name: Prometheus
    The name that will appear in Grafana.
  
  
  • 
    type: prometheus
    Specifies the data source type (Prometheus).
  
  
  • 
    access: proxy
    Grafana will proxy requests to Prometheus (recommended for most setups).
  
  
  • 
    url: http://prometheus:9090
    The URL where Prometheus is accessible from within the Docker network (using the service name).
  
  
  • 
    isDefault: true
    Sets this data source as the default in Grafana.
  
  
  
  
  
  

Summary:


This file ensures that Prometheus is pre-configured as the default data source in Grafana, so you don’t have to add it manually.

3. Start the Stack

docker-compose up -d

This command will pull the required images and start all services in the background.

4. Access the Services

• 
  Prometheus UI: http://localhost:9090
  


• 
  Grafana UI: http://localhost:3000
  
  
  
  
  • Default login: admin / admin
    
  
  
  
  
  
  



• 
  
  Node Exporter Metrics: http://localhost:9100/metrics
  

Prometheus UI showing that the, node exporter endpoint is connected successfully.

5. Add Dashboards in Grafana

• Prometheus is already set as a data source.


• Import community dashboards or create your own to visualize Node Exporter metrics.

• Click + Create Dashboard

• Import Node Exporter Dashboards: There are pre-configured dashboards available for Node Exporter. You can go to Grafana Dashboards and search for Node Exporter Full Dashboard (ID: 1860), which will give you a comprehensive view of your system metrics.

• In the Grafana Dashboards section, search for Node Exporter Full dashboard. The Dashboard ID for Node Exporter Full is 1860.


• Enter 1860 in the Import via ID field and click Load.


• In the Prometheus drop-down, choose the Prometheus data source you just configured.

• Click Import.


• Verify the Dashboard:


• Once the dashboard is imported, you should see a comprehensive view of your Ubuntu server’s metrics, including CPU usage, memory usage, disk I/O, network statistics, etc.

6. Stopping the Stack

docker-compose down

This will stop and remove all containers, but your data will persist in the named volumes.

Tips & Troubleshooting

• 
  Customizing Prometheus Targets: Edit prometheus.yml to add more scrape targets.


• 
  Persisting Data: Data is stored in Docker volumes (prometheus_data, grafana_data).


• 
  Logs: Check logs with:
  

  docker-compose logs prometheus

  docker-compose logs grafana

  docker-compose logs node_exporter

• 
  Change Default Password: For security, change the Grafana admin password in production.

Conclusion

With just a few files and Docker Compose, you can have a powerful monitoring stack up and running in minutes. This setup is perfect for local development, testing, or even small production environments.

Happy Monitoring!

References:

• Prometheus Documentation


• Grafana Documentation


• Node Exporter Documentation