Beta This project is experimental. Use at your own risk →

Always-on hardware that watches your network even when everything else is off.

A dedicated monitor is a device — Raspberry Pi, Mac Mini, or Docker container — running netglance 24/7 on your home network. It sees every device connect and disconnect, captures long-term trends, and alerts you to problems in real time. Unlike the background daemon on a laptop, it never sleeps.

When to use this mode

  • You want true 24/7 visibility, not just "when my laptop is open"
  • You care about catching events at 3 AM — rogue devices, DNS changes, certificate expirations
  • You want a historical record of your network state over weeks and months
  • You have a spare Raspberry Pi, Mac Mini, or a NAS/VM that can run Docker
  • You want a network monitoring appliance without buying commercial hardware

Choosing your platform

Platform Power draw Cost Setup time Best for
Raspberry Pi 4/5 ~5W ~$50-80 30 min Dedicated low-power appliance
Mac Mini ~6-40W (existing) 15 min Already have one sitting around
Docker varies free 10 min NAS, VM, cloud, or any Linux host

Raspberry Pi

The natural choice for a dedicated monitor. Low power draw means you can leave it plugged in forever. Ethernet gives stable, accurate readings. ARM-compatible Python ecosystem means netglance runs natively.

What you need: Pi 4 or 5, Ethernet cable, 32GB microSD (A2-rated), official power supply.

# After OS setup and SSH access:
pip install netglance
sudo netglance daemon install   # or use systemd — see full guide

Full Raspberry Pi setup guide → covers hardware selection, OS flashing, Python install, systemd service, log rotation, and GPIO status LED.

Mac Mini

If you have a Mac Mini that's always on (media server, home automation hub), add netglance as a launchd service. Same daemon setup as the Background Daemon mode, but on hardware that doesn't sleep.

uv tool install netglance
netglance daemon install

Full macOS daemon setup guide → covers launchd configuration, auto-restart, log rotation, and energy saver settings.

Docker

Run netglance in a container on any Linux host — NAS (Synology, QNAP), VM, cloud instance, or bare metal. Host networking is required for accurate network scanning.

docker run -d \
  --name netglance \
  --network host \
  --cap-add NET_RAW \
  --cap-add NET_ADMIN \
  -v netglance-data:/root/.config/netglance \
  netglance/netglance:latest \
  daemon start

Full Docker setup guide → covers Dockerfile, compose config, host networking, volume mounts, and container health checks.

What to expect

Once your dedicated monitor is running:

  • Device inventory updates every 15 minutes — new devices trigger alerts
  • DNS, TLS, and health checks run on schedule — failures trigger alerts
  • Baseline diffs capture network changes daily
  • Metrics accumulate in SQLite — query with netglance metrics for long-term trends
  • SSH in to run ad-hoc CLI commands or check netglance daemon status

Remote access

Your dedicated monitor runs headless. Access it via:

  • SSH: ssh pi@netglance.local (or by IP)
  • netglance REST API: netglance api serve --host 0.0.0.0 exposes results over HTTP for dashboards or other tools
  • MCP over HTTP: netglance mcp serve --transport http lets AI assistants query the monitor remotely

Pairing with other modes

  • Dedicated + MCP: Run the MCP server on your Pi/Mini so AI assistants can query the always-on monitor from any device on the network.
  • Dedicated + Scheduled Checks: Use cron for custom check combinations beyond the daemon's built-in schedule.
  • Dedicated + Alerts: Configure alert rules with notification channels (webhook, email) so you know the moment something changes, even if you're not at home.