2026-09-30

Why we test restores, not backups

Backup software is very good at reporting success. It tells you that a job finished, that a number of files were copied and that the destination accepted them. It does not tell you whether you can rebuild your office from those files on a bad Tuesday.

Our quarterly drills follow the same script: pick one system at random, restore it to isolated hardware, and time each step. The written report is short and boring on purpose. When something surprises us — a missing licence key, a database that needs a specific version, a password only one person knows — it goes into the recovery notes for next time.

If you only do one thing after reading this, schedule thirty minutes to restore a single folder from last month and see what happens.