It passed the tests. Now it's 3 AM and the pager is going off.

transcript

It passed the tests. You shipped it. Now it's 3 AM, and your pager is going off.

The alert: error rate over 5 percent. For 5 minutes straight. And you're the one on call. Is this a SEV1? Do you wake people up? Where do you even look?

A junior engineer says: I built it, it works on my machine. A mid-level engineer says: I built it, it passed the tests. A senior says: I can operate it. And I have a plan for when it breaks.

So, can you answer these? How do you know it's working? How do you deploy without breaking things? A bad deploy is the fastest way to an outage. How do you handle 10 times the traffic? And how do you stop it breaking again?

Five days. SLOs. Observability. Safe deploys. Capacity planning. Incidents. You've designed the systems. Now, can you run them? From I built it, to I can operate it.

Full breakdown in the lecture.