beginnerErrors & Observability
Errors Someone Can Act On
Decide what you would build from the brief alone. Everything below it is available, but the exercise stops working if you open it first.
The brief you were given
Your service returns 500 with a stack trace for everything that fails, including invalid input. Fix the error handling.
The trap — the wrong fix that looks right
Wrapping the whole application in one catch-all that logs and returns a generic 500. Errors stop leaking, which feels like the fix — and now every cause looks identical from outside. Validation failures inflate the 5xx rate until the alert threshold is meaningless, clients retry things that will never succeed, and the on-call engineer investigating an "error spike" finds it was a client sending a malformed date. The boundary is right; classifying before responding is the missing half.