VPS Monitoring

VPS Monitoring with Prometheus, Grafana & Node Exporter

This guide covers how to set up a full monitoring stack on a Linux VPS using Docker Compose. By the end, you will have Prometheus collecting metrics, Node Exporter exposing host-level system data, and Grafana displaying everything in a dashboard.

Architecture overview

Node Exporter (:9100) ──scrapes──> Prometheus (:9090) ──queries──> Grafana (:3000)

Node Exporter reads host metrics directly from the Linux kernel (/proc, /sys).
Prometheus scrapes those metrics on a fixed interval and stores them.
Grafana connects to Prometheus as a data source and renders dashboards.

Prerequisites

A Linux VPS with Docker and Docker Compose installed.
SSH access to the server (this guide uses Termius).
Basic familiarity with the terminal.

Project structure

Create the following folder structure on your VPS under /root/monitoring:

/root/monitoring/
├── docker-compose.yml
└── prometheus/
    └── prometheus.yml

Step 1 — Configure Prometheus

Create prometheus/prometheus.yml with a scrape job for each target:

global:
  scrape_interval: 5s

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

  - job_name: 'docker'
    static_configs:
      - targets: ['host.docker.internal:9323']

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

Note: Inside Docker, containers communicate by service name, not localhost. That is why the targets use names like prometheus and node-exporter instead of 127.0.0.1.

Step 2 — Configure Docker Compose

Create docker-compose.yml with three services — Prometheus, Grafana, and Node Exporter — all sharing the same monitoring network:

version: '3.8'

services:
  prometheus:
    image: prom/prometheus:latest
    container_name: prometheus
    restart: always
    ports:
      - "9090:9090"
    volumes:
      - ./prometheus/prometheus.yml:/etc/prometheus/prometheus.yml
    command:
      - "--config.file=/etc/prometheus/prometheus.yml"
    extra_hosts:
      - "host.docker.internal:host-gateway"
    networks:
      - monitoring

  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    restart: always
    ports:
      - "3000:3000"
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=admin
    volumes:
      - grafana-storage:/var/lib/grafana
    networks:
      - monitoring

  node-exporter:
    image: prom/node-exporter:latest
    container_name: node-exporter
    restart: always
    ports:
      - "9100:9100"
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - /:/rootfs:ro
    command:
      - "--path.procfs=/host/proc"
      - "--path.sysfs=/host/sys"
      - "--path.rootfs=/rootfs"
      - "--collector.filesystem.mount-points-exclude=^/(sys|proc|dev|host|etc)($$|/)"
    networks:
      - monitoring

networks:
  monitoring:

volumes:
  grafana-storage:

Key decisions explained

Setting	Why
`extra_hosts: host.docker.internal:host-gateway`	On Linux, `host.docker.internal` does not resolve automatically like on Mac/Windows. This line adds it manually so Prometheus can reach the Docker daemon metrics on port `9323`.
`/proc`, `/sys`, `/` mounted as `:ro`	Node Exporter needs to read kernel data from the host. The `:ro` flag makes the mounts read-only so the container can never write to the host filesystem.
`collector.filesystem.mount-points-exclude`	Filters out virtual kernel filesystems so disk usage metrics only reflect real drives.

Step 3 — Enable Docker daemon metrics

The docker job in Prometheus scrapes Docker engine metrics from port 9323. This endpoint is disabled by default on Linux and must be enabled manually.

On your VPS, edit /etc/docker/daemon.json:

nano /etc/docker/daemon.json

Add the following content:

{
  "metrics-addr": "0.0.0.0:9323",
  "experimental": true
}

Then restart the Docker daemon:

systemctl restart docker

Step 4 — Deploy to the VPS

Transfer the monitoring/ folder to your VPS using the Termius SFTP panel (drag and drop the folder to /root/monitoring), then start the stack:

cd /root/monitoring
docker compose up -d

Step 5 — Verify everything is running

Check that all containers are up:

docker compose ps

Check which ports are listening:

ss -tlnp

You should see ports 9090, 9100, 3000, and 9323 all listed.

Check Prometheus targets:

Open http://<your-vps-ip>:9090/targets in your browser. All three jobs — prometheus, docker, and node-exporter — should show status UP.

Check Node Exporter directly:

curl http://localhost:9100/metrics

You should see a large list of metrics prefixed with node_.

Step 6 — Connect Grafana to Prometheus

Open Grafana at http://<your-vps-ip>:3000 and log in with admin / admin.
Go to Connections → Data sources → Add data source.
Select Prometheus.
Set the URL to http://prometheus:9090.
Click Save & test — it should confirm the connection is working.

From here you can import a pre-built Node Exporter dashboard from grafana.com/dashboards (dashboard ID 1860 is a popular choice) or build your own.

Troubleshooting

host.docker.internal does not resolve

This happens on Linux when extra_hosts is missing from the Prometheus service. Add it to docker-compose.yml as shown in Step 2 and run docker compose up -d again.

Port 9323 connection refused

The Docker daemon metrics endpoint is not enabled. Follow Step 3 to edit daemon.json and restart Docker.

A target shows DOWN in Prometheus

Run docker compose ps to check all containers are running, then verify the service name in prometheus.yml matches exactly the service name in docker-compose.yml.

VPS Monitoring with Prometheus, Grafana & Node Exporter​

Architecture overview​

Prerequisites​

Project structure​

Step 1 — Configure Prometheus​

Step 2 — Configure Docker Compose​

Key decisions explained​

Step 3 — Enable Docker daemon metrics​

Step 4 — Deploy to the VPS​

Step 5 — Verify everything is running​

Step 6 — Connect Grafana to Prometheus​

Troubleshooting​