Mock Database Interviews

Introduction

These mock interviews simulate the structure of real backend / data engineering loops: a short live-coding pipeline, a system-design segment, and a postmortem question. Time each section — the constraint is part of the practice.

You'll find walkthrough answers for the first two; the others are intentionally left for you to write up and review with a peer.

Understanding the topic

3 timed interview rounds:

• R1 (30 min) — live aggregation: cohort retention by signup month.
• R2 (45 min) — system design: SaaS ticketing platform on MongoDB for 10k tenants.
• R3 (30 min) — postmortem: primary failed over twice in a week; describe your investigation.

Informative example

R1 walkthrough — cohort retention:

db.events.aggregate([
  { $match: { type:"login" } },
  // 1) attach signup month to every event
  { $lookup: { from:"users", localField:"userId", foreignField:"_id", as:"u" } },
  { $unwind: "$u" },
  { $set: { cohort: { $dateTrunc:{ date:"$u.createdAt", unit:"month" } },
            eventMonth: { $dateTrunc:{ date:"$ts", unit:"month" } } } },
  // 2) months since signup
  { $set: { monthOffset: { $dateDiff:{ startDate:"$cohort", endDate:"$eventMonth", unit:"month" } } } },
  // 3) unique active users per (cohort, offset)
  { $group: { _id:{ cohort:"$cohort", offset:"$monthOffset" }, users:{ $addToSet:"$userId" } } },
  { $project: { _id:0, cohort:"$_id.cohort", offset:"$_id.offset", active:{ $size:"$users" } } },
  { $sort: { cohort:1, offset:1 } }
]);

Best practices

• Always restate the question before coding — interviewers grade understanding first.
• Sketch the data model before writing the pipeline.
• Call out time/space trade-offs explicitly — even when not asked.