Skip to content
Interview rounds

Observability · Production incident

Three alerts, three services, four minutes

Several alerts fire at once across three services. How do you run the first ten minutes?

BackendSenior14 min
Interview-style scenarioLast reviewed: Checked against official documentation

Tuesday 19:42 · three services, two regionsIllustrative scenario, not a real incident

3 alerts in 4 minutes: checkout-api 5xx, cart-api p99, account-api 5xx

The shop runs in two regions, EU and US, each with its own copy of every service. Between 19:38 and 19:42 three alerts fired: errors on checkout, latency on cart and errors on account.

Each alert paged a different team. Three engineers have joined a call, and each is looking at their own service. The cart team released a new version 40 minutes ago.

What the dashboards and logs show

checkout-api 5xx
22% (usually 0.2%)
cart-api p99
4.8 s (usually 200 ms)
account-api 5xx
9% (usually 0.1%)
Releases today
cart-api at 19:02, nothing else

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"

A fifth of EU checkouts are failing and three engineers are on the call. 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.