Grafana Dashboard for Homelab: Build, Customize, and Monitor Everything

Reading time: ~14 minutes Audience: Homelabbers who want a single pane of glass for their infrastructure

What Is a Grafana Dashboard for Your Homelab?

Overview

Grafana is the open-source visualization platform that turns time-series data into beautiful, actionable dashboards. In a homelab, Grafana is the single pane of glass: CPU load, memory pressure, disk usage, network throughput, container health, and even UPS battery status—all in one place.

Why Grafana for homelabs: - Free and open source: AGPLv3. No feature limits. - Data source agnostic: Prometheus, InfluxDB, Loki, MySQL, PostgreSQL, Zabbix, and 100+ others. - Community dashboards: Import pre-built dashboards for Node Exporter, cAdvisor, Proxmox, Pi-hole, and more. - Alerting: Native alerting rules with email, Slack, Telegram, and webhook support. - Mobile-friendly: Responsive web UI works on tablets and phones.

Key Benefits

1. Proactive maintenance: See disk fill-up before it happens
2. Performance tuning: Identify which VM is consuming RAM or IOPS
3. Capacity planning: Track 30-day trends to know when to upgrade
4. Troubleshooting: Correlate events across multiple systems in one timeline
5. Show off: A well-built Grafana dashboard is the homelab equivalent of a trophy case

Prerequisites

Hardware Requirements

Software Requirements

• Docker and Docker Compose
• A data source (Prometheus, InfluxDB, or both)
• Node Exporter and cAdvisor running on monitored hosts

Knowledge Prerequisites

• Basic Prometheus query language (PromQL) or InfluxQL
• Understanding of time-series data
• Familiarity with Docker networking

Step 1: Deploy Grafana with Docker Compose

Objective

Run Grafana as a container with persistent storage and pre-configured data sources.

Step-by-Step Instructions

1. Create a directory for Grafana:

mkdir -p ~/grafana/config
cd ~/grafana

1. Create docker-compose.yml:

services:
  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    restart: unless-stopped
    ports:
      - "3000:3000"
    environment:
      - GF_SECURITY_ADMIN_USER=admin
      - GF_SECURITY_ADMIN_PASSWORD=***      - GF_USERS_ALLOW_SIGN_UP=false
      - GF_SERVER_ROOT_URL=https://grafana.yourdomain.com
      - GF_INSTALL_PLUGINS=grafana-clock-panel,grafana-simple-json-datasource
    volumes:
      - grafana-data:/var/lib/grafana
      - ./config/datasources.yml:/etc/grafana/provisioning/datasources/datasources.yml
      - ./config/dashboards.yml:/etc/grafana/provisioning/dashboards/dashboards.yml
      - ./dashboards:/var/lib/grafana/dashboards

volumes:
  grafana-data:

1. Create config/datasources.yml to auto-provision Prometheus:

apiVersion: 1

datasources:
  - name: Prometheus
    type: prometheus
    access: proxy
    url: http://prometheus:9090
    isDefault: true
    editable: false

  - name: InfluxDB
    type: influxdb
    access: proxy
    url: http://influxdb:8086
    database: homelab
    editable: false

1. Create config/dashboards.yml to auto-provision dashboard files:

apiVersion: 1

providers:
  - name: 'homelab'
    orgId: 1
    folder: 'Homelab'
    type: file
    disableDeletion: false
    updateIntervalSeconds: 10
    allowUiUpdates: true
    options:
      path: /var/lib/grafana/dashboards

1. Start the container:

docker compose up -d

1. Access Grafana at http://your-server-ip:3000 (login: admin / password from env)

Step 2: Build the “Homelab Overview” Dashboard

Objective

Create a dashboard with panels for system metrics, Docker containers, and network.

Step-by-Step Instructions

1. In Grafana, click + > New Dashboard > Add visualization

2. Panel 1: CPU Usage (Stat)

3. Data source: Prometheus
4. Query: 100 - (avg by(instance) (irate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)
5. Visualization: Stat
6. Thresholds: Green < 60, Yellow < 80, Red > 80

7. Title: “CPU Usage”

8. Panel 2: Memory Usage (Gauge)

9. Data source: Prometheus
10. Query: 100 * (1 - (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes))
11. Visualization: Gauge
12. Thresholds: Green < 70, Yellow < 85, Red > 85

13. Title: “Memory Usage”

14. Panel 3: Disk Usage (Table)

15. Data source: Prometheus
16. Query: 100 * (1 - node_filesystem_avail_bytes{fstype!="tmpfs"} / node_filesystem_size_bytes)
17. Visualization: Table

18. Title: “Disk Usage by Mount”

19. Panel 4: Network Traffic (Time Series)

20. Data source: Prometheus
21. Query A (RX): irate(node_network_receive_bytes_total{device!="lo"}[5m])
22. Query B (TX): irate(node_network_transmit_bytes_total{device!="lo"}[5m])
23. Visualization: Time Series
24. Unit: bytes/sec (IEC)

25. Title: “Network Traffic”

26. Panel 5: Top 5 Containers by CPU (Bar Chart)

27. Data source: Prometheus
28. Query: topk(5, rate(container_cpu_usage_seconds_total{name!=""}[5m]))
29. Visualization: Bar Chart

30. Title: “Top Containers by CPU”

31. Panel 6: Docker Container Health (Stat)

32. Data source: Prometheus
33. Query: count(container_last_seen{name!=""})
34. Visualization: Stat

35. Title: “Active Containers”

36. Save the dashboard as homelab-overview.json in the dashboards/ folder. Grafana will auto-load it.

Step 3: Import Community Dashboards

Objective

Import battle-tested dashboards for common homelab tools.

Step-by-Step Instructions

1. In Grafana, click + > Import Dashboard

2. Enter the dashboard ID and click Load:

1. Select your Prometheus data source
2. Click Import

Step 4: Customize Panels and Variables

Objective

Add template variables so one dashboard serves multiple hosts.

Step-by-Step Instructions

1. In your dashboard, click Settings > Variables > New Variable

2. Create a variable named node:

3. Data source: Prometheus
4. Query: label_values(node_uname_info, nodename)
5. Refresh: On Dashboard Load
6. Multi-value: Yes

7. Include All option: Yes

8. Update your queries to use the variable:

# Before
100 - (avg by(instance) (irate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

# After
100 - (avg by(instance) (irate(node_cpu_seconds_total{mode="idle", nodename=~"$node"}[5m])) * 100)

1. Now you can select specific hosts or “All” from a dropdown at the top of the dashboard.

Step 5: Set Up Alerts in Grafana

Objective

Create alerting rules directly in Grafana (separate from Alertmanager).

Step-by-Step Instructions

1. Go to Alerting > Alert Rules > New Alert Rule

2. Create a rule for disk space:

3. Query: 100 * (1 - node_filesystem_avail_bytes / node_filesystem_size_bytes) > 90
4. Evaluation: Every 1m for 5m
5. Labels: severity=critical

6. Contact point: Telegram or Slack

7. Create a rule for high memory:

8. Query: 100 * (1 - node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) > 85
9. Evaluation: Every 1m for 5m

10. Labels: severity=warning

11. Test the alert by temporarily lowering the threshold.

Pro Tips

Tip 1: Use Row Panels for Organization

Group related panels into collapsible rows. Common rows for a homelab dashboard: - Overview: CPU, memory, disk, uptime - Network: Traffic, errors, top talkers - Containers: CPU, memory, restart count - Storage: ZFS pool health, disk temperatures, IOPS - Services: Per-service panels (Nextcloud, Plex, Jellyfin)

Tip 2: Add Annotations for Events

Grafana annotations mark events on the timeline. Use the API to add annotations for deployments, reboots, or backups:

curl -X POST http://grafana:3000/api/annotations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "dashboardId": 1,
    "panelId": 1,
    "time": '"$(date +%s)000"',
    "tags": ["deployment", "nextcloud"],
    "text": "Deployed Nextcloud 30.0.1"
  }'

Tip 3: Export and Version Control Your Dashboards

Dashboard JSON should be in Git. Export via Share > Export > Save to file. Commit to a repo and use Grafana provisioning to auto-deploy. This prevents losing work and makes migrations trivial.

Troubleshooting Common Issues

Problem 1: “No Data” in Panels

Check: Ensure Prometheus is scraping the target. Go to Explore in Grafana and run the query manually. If no data, check the target in Prometheus (http://prometheus:9090/targets).

Fix: Common causes: wrong job name, container not on the same Docker network, or firewall blocking port 9100.

Problem 2: Dashboard Variables Not Working

Check: Verify the variable query returns values in Explore.

Fix: If the variable is empty, the label name may be wrong. Use label_values(node_uname_info) to see all available labels.

Problem 3: Slow Dashboard Loading

Check: Complex queries with high cardinality (many unique label combinations) can slow Grafana.

Fix: Add aggregation. Instead of container_cpu_usage_seconds_total, use sum by(name) (container_cpu_usage_seconds_total). Limit time ranges. Use recording rules in Prometheus for expensive queries.

Conclusion

Summary

Grafana is the visualization layer that makes your Prometheus data actionable. A well-built homelab dashboard gives you proactive insight into CPU, memory, disk, network, and container health. Community dashboards provide a fast starting point; custom panels and variables tailor the experience to your infrastructure.

Next Steps

1. Import the Node Exporter Full dashboard (ID 1860)
2. Build a custom “Homelab Overview” dashboard with your top 6 metrics
3. Add template variables for multi-host monitoring
4. Set up Grafana alerting for disk space and memory pressure
5. Export your dashboards to JSON and commit them to Git

Affiliate Opportunities

• Mini PCs for Grafana: Minisforum, Beelink (Amazon)
• Displays: Wall-mounted tablets for dashboard kiosks (Amazon)
• VPS: Hetzner for offsite Grafana (referral)

Internal Linking Strategy

• prometheus-setup → guide: “prometheus-monitoring-homelab.md”
• alertmanager → guide: “prometheus-alertmanager-setup.md”
• docker-monitoring → guide: “docker-monitoring-grafana-prometheus.md”
• node-exporter → guide: “prometheus-monitoring-homelab.md”

CTA

• [comment] Share a screenshot of your Grafana dashboard! What panels are essential for your homelab?
• [newsletter] Subscribe for our dashboard JSON packs and monitoring stack updates.
• [internal_link] Need to set up Prometheus first? Read our Prometheus monitoring guide next.