How to Set Up Prometheus, InfluxDB, and Grafana for Full-Stack Monitoring with Docker

Monitoring your infrastructure is critical to keeping systems healthy, optimizing performance, and planning for growth. In this guide, you’ll learn how to set up a powerful monitoring stack using Prometheus and InfluxDB to collect and store metrics, and Grafana to visualize them.

To demonstrate how data collection works, we’ll use two popular Prometheus exporters: cAdvisor for container metrics and Node Exporter for system-level data. All components will be combined into a single docker-compose.yml file and run using Docker Compose on an Ubuntu server.

🔧 What You’ll Set Up:

• Core Monitoring Stack:
  • Prometheus: Collects and stores metrics from various sources.
  • InfluxDB: Optional storage for custom or long-term time-series data.
  • Grafana: Visualizes metrics from Prometheus and InfluxDB.
• Demonstration Exporters:
  • cAdvisor: Provides Docker container metrics to Prometheus.
  • Node Exporter: Provides Linux system-level metrics like CPU, memory, and disk usage to Prometheus.
• Service Management:
  • Docker Compose: Simplifies running and managing all services.

📂 Project Directory Structure

grafana/
├── data/
│   ├── grafana/
│   ├── influxdb2/
│   │   └── env/
│   └── prometheus/
│       ├── config/
│       │   └── prometheus.yml
│       └── data/
├── docker-compose.yml

📌 Prerequisites

Make sure the following are in place:

• Ubuntu 20.04 or higher
• Docker & Docker Compose installed
• sudo privileges

Step 1: Prepare Directories and Permissions

Let’s create the necessary directories and assign the correct permissions:

sudo mkdir -p grafana/data/grafana grafana/data/prometheus grafana/data/influxdb2 && cd grafana
sudo chown -R 472:472 ./data

📌 Note: UID 472 is the default user Grafana runs as in its container. Assigning this ownership ensures Grafana can write to the volumes.

Step 2: Set Up InfluxDB Admin Credentials

sudo mkdir -p ./data/influxdb2/env && \
sudo bash -c 'echo "Your_InfluxDB_Admin_User" > ./data/influxdb2/env/.env.influxdb2-admin-username' && \
sudo bash -c 'echo "Your_InfluxDB_Admin_Password" > ./data/influxdb2/env/.env.influxdb2-admin-password' && \
sudo bash -c 'echo "MyInitialAdminToken0==" > ./data/influxdb2/env/.env.influxdb2-admin-token'

• Replace the placeholders:
  • Your_InfluxDB_Admin_Username – Replace with your desired InfluxDB admin username.
  • Your_InfluxDB_Admin_Password – Replace with your preferred secure password.
  • MyInitialAdminToken0== – Replace with your custom admin token, or leave it as is if you don’t want to set a specific token.

🔐 Security Tip: Make sure these files are not exposed to the public and are only readable by trusted processes. You can restrict access with:

sudo chmod 600 ./data/influxdb2/env/.env.influxdb2-*

Step 3: Configure Prometheus

Create the Prometheus configuration file:

sudo nano ./data/prometheus/config/prometheus.yml

Add the following content inside prometheus.yml:

global:
  scrape_interval: 15s

scrape_configs:
  - job_name: 'prometheus'
    static_configs:
      - targets: ['localhost:9090']

  - job_name: 'node-exporter'
    static_configs:
      - targets: ['node-exporter:9100']

  - job_name: 'cadvisor'
    static_configs:
      - targets: ['cadvisor:8080']

This configuration tells Prometheus where to scrape metrics from, including itself, Node Exporter, and cAdvisor.

Step 4: Create docker-compose.yml