I’ve decided to take a leaf out of my professional life and apply that to some of my homelab life (because it is life!).

So here I am doing some public incident management of stuff that should have little to any affect on people outside my home.

This is made up of a couple of pieces:

  1. Recording my incidents with rigor
  2. Advertising the public facing things

To do option 1, I cheated. I used to Claude Code to grab a bunch of information, even perform a bunch of actions, collate it all together, process it with some rules and create a markdown file with the analysis.

To do option 2, I cheated. I used Claude Code to take the status from my home Nagios monitoring system, pick out some important hosts and services and display that status on a public website.

Option 1 - Incident docs

To record and display this, I’m keeping the markdown docs in a repo and using mkdocs to generate html files and serve on a my Incidents site (hosted by GitHub Pages).

Option 2 - Public Status page

I created and open sourced a python API to grab the status of some carefully selected Nagios hosts and services. Then, I grab this with a vue.js front end to make it look pretty.

Here’re my hosts and services I advertise

I preseent to you, the very aptly named: nagios-public-status-page

PS

If the Statuspage site is unavailable, that’ll be because my home internet is borked, or my k8s cluster isn’t clustering very well.