Notes
- Backups: a backup you have never restored is a hope, not a backup. Restore into a scratch directory once a quarter.
- Upgrades: change one thing at a time. If two things changed and something broke, you now have two suspects.
- Disk space: check logs and container layers first; they grow quietly and fail loudly.
- Time: keep clocks in sync. Certificates, tokens and logs all depend on it.
- Passwords and keys: keep them out of notes, scripts and shell history; use a password manager and key files with tight permissions.
- Documentation: the best time to write down a fix is right after you made it.