Observability · Production incident
One country says it is slow. Every dashboard is green
Users in one region report slow responses while the global dashboards show nothing. How do you find the cause and what do you do?
Tuesday 11:50 UTC · api-gateway, three regionsIllustrative scenario, not a real incident
Support: 40 tickets in an hour saying the app is slow, all from one country. No alert has fired
The API is served from three regions, and each user is routed to the nearest one. Support has forty tickets since 11:00 UTC, all from one country, where it is early evening.
The latency dashboard shows a global p99 of 195 ms against an SLO of 300 ms, and the error rate is flat. Nothing was released today.
What the dashboards and logs show
- Global p99 latency
- 195 ms, SLO 300 ms
- Global error rate
- 0.02%, normal
- Support tickets
- 40 in an hour, one country
- Last release
- yesterday 14:00 UTC, all regions
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"
You have the tickets and a green dashboard. 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.