Stop 15 · Interview PreparationProcess & Strategy · 3/3

Performance Testing Scenarios

Part of the QA Testing Roadmap

In short

'How would you test if this system can handle 10,000 users?' is a favourite scenario. A good answer clarifies what '10,000 users' means, picks the right test type, and names the metrics.

'How would you test if this system can handle 10,000 concurrent users?' — testing Phase 10's performance testing knowledge applied to a concrete scenario.

The purple blob holds a stopwatch beside a ticket booth as a big crowd rushes towards it.

Think of it like this

Tatkal booking at 10 a.m.

Railway Tatkal booking opens at a fixed time, and a huge crowd hits the site in the same minute. The site is fine all day, then gets hammered. Performance scenarios ask: what happens at that exact moment?

Watch it happen

Pick the right type of performance test

Perf tests
Load
Stress
Spike
Soak
Expected peak
Beyond the limit
Sudden jump
Many hours

1/5Different questions need different tests. Name the right one in your answer.

The big ideas

What you need to know

Clarify the number

10,000 at the same moment, or per hour? Logged in, or browsing? The answer changes the whole test.

Realistic flows

Model what users really do: browse, search, add to cart, pay, with think time between steps.

The right metrics

Response time percentiles like p95, error rate, throughput, and server CPU, memory and database load.

Like-for-like environment

Results only mean something on an environment sized like production, with permission to load it.

How it works

A strong answer, step by step

  1. 1

    Clarify

    Concurrent or per hour, which user flows, and what response time is acceptable.

  2. 2

    Choose test types

    Load test for the target, then a spike or stress test to find limits.

  3. 3

    Script and ramp

    Script real user flows in k6 or JMeter and ramp users gradually.

  4. 4

    Measure and report

    Watch p95, errors and server metrics, then name the bottleneck and next step.

Save yourself the pain

Common mistakes

Oops

Answering only with a tool name: 'I'd use JMeter'.

Do this instead

Tools come last. Start with clarifying questions, test type, user flows and metrics.

Oops

Running the load test from your laptop against production without telling anyone.

Do this instead

Use an agreed environment and time window. Unannounced load can take down a live service.

Your mission

Answer three scenarios

~25 min

Quick check

Users report the app gets slower the longer the servers run, then needs a restart. Which test fits best?

Go deeper

  1. 1Review Phase 10 and practice designing a load test plan for a sample scenario.
  2. 2Prepare to discuss key performance metrics and how you'd interpret results.
  3. 3Practice explaining the difference between load, stress, and spike testing clearly.

More process & strategy at this stop

Stuck on this topic? Ask an Insider

Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.

Download
InsideEdge