Observability · Production incident
All green, and the customers cannot pay
Dashboards are green but customers report errors. What did you fail to measure?
Tuesday 16:20 · checkoutIllustrative scenario, not a real incident
Support has 140 tickets about failed checkouts in one country, and no alert has fired
Every panel on the checkout dashboard is green: server error rate, median latency, health checks. The on-call engineer has looked twice and found nothing.
The tickets are real. They come from two groups: customers in one country, and people on the newest version of the mobile app.
What the dashboards and logs show
- Server 5xx rate
- 0.02%, normal
- Health checks
- passing on every instance
- Median latency
- 80 ms, normal
Clarify · optional · scored under "Hypothesis before change"
Ask the interviewer
0 of 3 asked
Each question reveals a fact. What you ask before you act is part of your score.
First move · scored under "Mitigate before root cause"
Customers are failing now and no graph shows it. What do you do first?
Your rubric score, the model answer, the follow-up questions and the related lessons appear once you have answered the last step.