Overnight, Cascade Cloud had a brief outage —read the status pageCascade Cloud is a fictional cloud platform used for this demo: independent regions, managed databases, and a generous free tier.— and the on-call team rerouted traffic within minutes. It's usually rock solid, but nothing stays up forever.

The rollback itself was handled by Priya Nandan

, who has been on call for three years straight and still answers pages faster than the dashboards load.

Afterwards the team wrote up a short retro.

“Nobody panicked. We rerouted, rolled back, and were back to normal before the coffee finished brewing.”

It's now the template every incident review starts from.