Notes on small, durable systems
I run a handful of servers for myself, family and a couple of small clients. These are the notes I wish I had found the first time: what broke, what I changed, and what has quietly kept working since.
-
The restore drill: testing backups you have never restoredBackups are a hypothesis until you restore one. A quarterly routine with restic and a scratch VM.
-
nginx on a 512 MB boxWorker counts, buffer sizes and the three settings that actually mattered on a tiny VPS.
-
Moving my cron jobs to systemd timersLogging, missed runs and randomized delays — why I finally switched, and the one thing I miss.
Older notes from before the move to this domain are being cleaned up and will return gradually.