This week’s goal
Patch the server without turning a normal evening into accidental incident response.
I wanted updates to feel routine, not like I had just pulled a random wire out of a wall.
What I actually did
I created a small update checklist:
- check backups first
- note what I am about to update
- update one layer at a time
- restart services deliberately
- confirm Jellyfin still works
- check the server after a reboot
This is not exciting. That is the point.
What broke, or nearly did
I almost updated everything at once because the command was right there and I briefly felt powerful.
This is how trouble starts. Not with malice. With confidence and a terminal prompt.
What I’m keeping
- Update notes
- Backup check before changes
- Reboot test after bigger changes
- No late-night “quick updates”
The last rule is important because “quick updates” after 10pm are just elaborate traps wearing a sensible hat.
SmallGrid takeaway
Updates are safer when they have a routine. A small checklist beats heroic troubleshooting after you have forgotten what you changed.