Incident response

Production incident: p95 latency spike

It's 2:14 PM on a Tuesday. You're the on-call engineer for an online store. Your phone buzzes: "checkout-api p95 latency 2.4 s (normal 180 ms), error rate 3% (normal 0.4%)." Customers are paying for orders right now.

About 6 minutes · 10 decision points possible · every choice gets instant feedback