Grafana: Monitoring for Homelab

The Blind Flight Problem

Imagine you run a homelab with ten, twenty, or thirty services. Everything seems to be running fine. Then one day a customer tells you your website has been offline for six hours. Or you realize your backup disk has been full for two weeks. Or a Docker container has been silently restarting every fifteen minutes for hours — and nobody noticed.

That happens without monitoring. You are flying blind. Monitoring is the difference between "finding problems before anyone is affected" and "learning about problems from angry users."

What Good Monitoring Does

• Shows current state: Is everything healthy right now? Which services are running? How much disk space is left? You should be able to answer these in under ten seconds.
• Shows trends over time: Your disk was at sixty percent last month and is now at eighty. At this rate, it will be full in six weeks. Trends tell you what is coming.
• Alerts you to problems: When a threshold is exceeded — disk over ninety percent, service down for more than two minutes — monitoring sends you a notification in seconds, not hours.

Grafana is the dashboard tool for everything with metrics. Whether Docker containers, server resources or your own applications - Grafana visualizes everything.

Why Grafana?

• Open Source - free
• 100+ data sources - Prometheus, InfluxDB, Elasticsearch
• Flexible dashboards - custom visualizations
• Alerting - notifications when problems occur
• Self-hosted - all data stays local

Components

Installation

services:
  prometheus:
    image: prom/prometheus
    ports:
      - "9090:9090"
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml

  grafana:
    image: grafana/grafana
    ports:
      - "3001:3000"
    volumes:
      - ./grafana-data:/var/lib/grafana

Key Metrics

• CPU: usage_percent, load
• Memory: used_bytes, available_bytes
• Disk: read_bytes, write_bytes
• Network: rx_bytes, tx_bytes
• Docker: container_status, health