Automation that survives reboots: notes from Raspberry Pi work

Production is not your laptop

On a desk you run a script and watch logs. In the field or a rack cabinet you need repeatable boot, log rotation, and a little observability.

Habits that pay back

• A real service unit instead of manual nohup: start on boot, restart on failure.
• Configuration outside the code: paths, tokens, endpoints (never commit secrets).
• The reboot test: if it fails after power cycle, you want to know before go-live.

If this sounds familiar

Send a short note via the home page contact section with the data source and the outcome you need. We will propose a lean path, not a slide deck.