Six Grafana dashboards you generate instead of draw, alert rules short enough that you are still reading them in six months, and a liveness guard where green means the job actually ran.
$49 minimum, $59 suggested
Pay what you think it's worth, with $49 as the floor. Team licence (5 seats) $169.
Available September 15. Join the list to get it first.
Version 1.0, six dashboards, a generator, a liveness guard, 25 alert rules, a 20-page runbook and two worksheets, updated September 2026.
Free updates forever. 30-day 100% refund, no questions.
No subscription, no account, no DRM.
Most homelab monitoring ends up in one of two states. Either you have four dashboards you copied from somewhere, three of which show "No data" and none of which you trust — or you have forty alert rules firing continuously and you have quietly stopped reading the notifications. This is what I run instead.
build.py --check fails
when the JSON has drifted from its source, so hand-edits get caught.unmeasured, never a confident zero, because "we measured nothing" and "the number is
nothing" must not look the same.repeat_interval, and the one rule that keeps a list readable: no
alert without an action.Dashboards are generated, not drawn. Six hand-maintained JSON files drift. The generator enforces the house rules with assertions: every panel has a description, no panel hardcodes a datasource, no two panels overlap.
Green means it ran. This came out of a real failure — a status tile that was green for nine days while the machine running the job was off. Nothing alerted, because nothing had failed. Things had simply stopped happening.
${DS} datasource variable, so they import into any Grafana and work.lib.py style system, build.py, and a
deploy.sh that pushes to the Grafana API from environment variables.liveness.py and a one-line
liveness.sh wrapper you put in front of a cron line, plus systemd units and a cron
example. Records last run, last success, exit code and duration; exports Prometheus textfile
metrics and a JSON status.thresholds.md — one line per rule on why that number and not
another.denoising.md — dedupe, inhibition, severity routing,
repeat_interval, and a working alertmanager.yml.job label trap, the cluster-exporter
multiplication trap.Everything is plain text: Python, YAML, JSON, Markdown, one PDF. Python standard library only, no
pip install. No app, no account, nothing that stops working if this site does.
Vendor-neutral. The Proxmox, container and endpoint boards need their exporters; without them
those panels read unmeasured rather than breaking.
No. Prometheus, Grafana and node_exporter are the only requirements. Nothing in the kit is tied to my hardware — no hostnames, no IPs, no tokens.
The JSON is the smallest part. What you are buying is the generator that keeps six boards consistent as you change them, the liveness guard that catches jobs that silently stopped, and the alert rules with the reasoning for each threshold written next to it. Free dashboards give you panels; this gives you a system you can still maintain in a year.
No. Free updates forever. The same download link always serves the newest version, and you get an email when one ships.
Yes — 30 days, 100%, no questions asked. Email me and I'll refund it. You keep the files.
The dashboards, the generator and the liveness guard are what I run in my own lab, and the nine-day green tile was my own. I used AI as an editor to tighten the prose.
Not yet. I'd rather show none than invent them. The first buyers get asked for a one-line review, and those go here when they exist.
Not ready to buy? Take the 96-item Homelab Ops Checklist, which costs nothing, or read the guides.
Made by a working sysadmin who runs this lab. About Mark Restivo.