Run it also means keep it running. Today, a short one, we open systems admin, the quiet job of keeping your machines up, fed, and backed up. By the end you will know the few moves that stop a small problem from becoming a dead website at the worst possible hour.

The first move is to watch the machine before it breaks. Set a simple check that pings your server every minute and tells you if it stops answering. You can wire that check into your agent so it reads the status and warns you in plain words. A machine that is about to fall over usually says so first.

The second move is backups you have actually tested. Schedule a copy of your data to separate storage on a cron clock. Then, and this is the part people skip, restore it once to prove it works. A backup you never restored is just a hope with a file size.

The third move is a runbook, a short page that says: if the site is down, do this, then this. Keep it where your agent can read it. When something breaks at two in the morning, you do not want to think. You want to follow the list you wrote when you were calm.

The proof is in our own house. The studio keeps these runbooks for its own systems, the same plain checklists we just described, so a recovery is a few known steps instead of a panic. We eat the cooking before we serve it.

So systems admin is not glamorous and that is the point. Watch the numbers, test your backups, write the runbook. Do the dull work now so the loud emergency never arrives. Tomorrow we lock the doors. See how we run our own at office.temerarii.xyz.