Introduction

You have a low-cost VPS running a few Docker containers and maybe a website. Everything seems fine — until:

• You can’t SSH in because the disk is full of logs
• Your site returns 502 errors, and the third-party monitoring service just expired
• A memory leak OOM-kills your VPS, and you have no idea when it happened

Don’t leave monitoring to intuition. Today we’ll set up a complete monitoring system on your VPS using three open-source tools that cost exactly $0:

The whole stack uses just 200MB of RAM — any 1GB VPS can handle it with room to spare.

1. Architecture Overview

┌─────────────────────────────────────┐
│          Your VPS                    │
│  ┌──────────┐  ┌───────────────┐    │
│  │  Node     │  │  Prometheus   │    │
│  │  Exporter │──│  (scrape)     │    │
│  │  :9100    │  │  :9090       │    │
│  └──────────┘  └───────┬───────┘    │
│                        │            │
│                 ┌──────▼───────┐    │
│                 │   Grafana    │    │
│                 │  :3000       │    │
│                 └──────┬───────┘    │
│                        │            │
│            ┌───────────▼──────┐     │
│            │  Browser / Phone │     │
│            └──────────────────┘     │
└─────────────────────────────────────┘

• Node Exporter collects CPU, memory, disk, and network metrics, exposing them on :9100
• Prometheus periodically scrapes Node Exporter’s data and stores it in a time-series database
• Grafana reads from Prometheus to render dashboards and trigger alerts

2. One-Click Deploy with Docker Compose

Create a project directory:

mkdir -p ~/monitoring && cd ~/monitoring

Create docker-compose.yml:

version: '3.8'

services:
  prometheus:
    image: prom/prometheus:v2.53.0
    container_name: prometheus
    restart: unless-stopped
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml
      - prometheus_data:/prometheus
    ports:
      - "9090:9090"
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--storage.tsdb.retention.time=30d'

  node_exporter:
    image: prom/node-exporter:v1.8.0
    container_name: node_exporter
    restart: unless-stopped
    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)($$|/)'
    ports:
      - "9100:9100"

  grafana:
    image: grafana/grafana:11.1.0
    container_name: grafana
    restart: unless-stopped
    volumes:
      - grafana_data:/var/lib/grafana
    ports:
      - "3000:3000"
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=admin   # ← Change this immediately after first login!
      - GF_INSTALL_PLUGINS=grafana-piechart-panel

volumes:
  prometheus_data:
  grafana_data:

Create the Prometheus config file prometheus.yml:

global:
  scrape_interval: 15s
  evaluation_interval: 15s

alerting:
  alertmanagers: []

rule_files: []

scrape_configs:
  - job_name: 'node'
    static_configs:
      - targets: ['node_exporter:9100']

Start everything:

docker compose up -d

Verify everything is running:

# Check Prometheus targets
curl -s http://localhost:9090/api/v1/targets | jq .

# Check Node Exporter metrics
curl -s http://localhost:9100/metrics | head -20

# Access Grafana
echo "Open http://YOUR_VPS_IP:3000  (admin / admin)"

3. Configure the Data Source in Grafana

1. Open http://YOUR_VPS_IP:3000 in your browser
2. First login: username admin, password admin (you’ll be prompted to change it)
3. Go to Connections → Data Sources → Add data source
4. Select Prometheus
5. Set the URL to http://prometheus:9090
6. Click Save & Test — you should see a green success message

4. Import the Classic Dashboard

Grafana dashboard ID 1860 (Node Exporter Full) gives you a beautiful, comprehensive dashboard:

# Option 1: Import via Grafana CLI
docker exec grafana grafana cli \
  plugins install grafana-piechart-panel

# Option 2: Import via Web UI
# 1. Click "+" → Import on the left sidebar
# 2. Enter ID: 1860
# 3. Select your Prometheus data source
# 4. Click Import

After importing, you’ll see:

• 📊 CPU Usage — real-time load per core
• 💾 Memory Usage — RAM, Swap, cache distribution
• 💽 Disk Space — partition usage, IO read/write
• 🌐 Network Traffic — inbound/outbound graphs
• ⏱ System Uptime — and Load Average

5. Configure Email Alerts (Optional but Recommended)

5.1 Set up SMTP in Grafana

Add SMTP configuration to the Grafana environment variables in docker-compose.yml:

environment:
  - GF_SMTP_ENABLED=true
  - GF_SMTP_HOST=smtp.gmail.com:587
  - GF_SMTP_USER=youraccount@gmail.com
  - GF_SMTP_PASSWORD=your-app-specific-password
  - GF_SMTP_FROM_ADDRESS=youraccount@gmail.com
  - GF_SMTP_FROM_NAME=VPS Alert

Restart Grafana:

docker compose up -d grafana

5.2 Create Alert Rules

In Grafana:

1. Go to Alerting → Alert rules → New alert rule
2. Set condition: disk_usage_percent > 85
3. Set evaluation interval: every 5 minutes
4. Add a contact point with your email
5. Save and enable

Recommended alert rules:

6. Security Hardening

6.1 Nginx Reverse Proxy with HTTPS

# /etc/nginx/sites-available/grafana
server {
    listen 443 ssl;
    server_name grafana.yourdomain.com;

    ssl_certificate /etc/letsencrypt/live/yourdomain.com/fullchain.pem;
    ssl_certificate_key /etc/letsencrypt/live/yourdomain.com/privkey.pem;

    location / {
        proxy_pass http://127.0.0.1:3000;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
    }
}

6.2 Restrict Port Exposure

Expose only Grafana (via reverse proxy). Keep Prometheus and Node Exporter on the internal Docker network:

# In docker-compose.yml, remove ports: from Prometheus and Node Exporter
# Use expose: instead for internal-only communication
services:
  prometheus:
    expose:
      - "9090"    # internal network only
    # remove ports: section

  node_exporter:
    expose:
      - "9100"    # internal network only
    # remove ports: section

6.3 Set a Strong Grafana Password

# Use a strong password via environment variable
environment:
  - GF_SECURITY_ADMIN_PASSWORD=Your_Strong_P@ssw0rd!
  - GF_SECURITY_DISABLE_INIT_ADMIN_CREATION=false

7. Real-World Resource Usage

Tested on a 1GB RAM / 1 vCPU / 20GB disk budget VPS:

Bottom line: A 1GB VPS is more than enough. A 30-day retention period uses ~10GB of disk — plan for 15GB+ to be safe.

8. Going Further: More Exporters

This stack is easily extensible. Add whatever exporter matches your needs:

Just add the corresponding job to Prometheus’s scrape_configs and they’ll automatically be included.

Summary

Prometheus + Grafana + Node Exporter is the “golden trio” of VPS monitoring — completely free, backed by massive communities, and highly extensible. Compared to paid third-party services like Datadog or New Relic, this self-hosted solution runs smoothly on a budget 1GB VPS and keeps your data entirely under your control.

Key takeaways:

• ✅ One-click deploy with Docker Compose, up in 5 minutes
• ✅ Classic dashboard ID 1860 — import and go
• ✅ Email alerts configured in minutes
• ✅ Nginx reverse proxy + HTTPS keeps access secure
• ✅ Extensible to containers, databases, and external sites

Remember: Monitoring is not optional — it’s the foundation of reliable operations. Set this up before your next outage, not after.