Concept questionMedium~7 minTesting & Quality / 9 of 20

Interview question

How would you integration-test an ASP.NET Core API with `WebApplicationFactory`?

Integration-test ASP.NET Core with WebApplicationFactory

Uses the real ASP.NET Core pipeline while replacing only environment-specific boundaries deliberately.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

I derive a custom WebApplicationFactory<Program>, create an HttpClient, and call the API over HTTP. The test boots the real service registration and request pipeline with test configuration. I replace external services and the database only through explicit test registrations, then assert status code, headers, and response contract rather than calling the endpoint method directly.

Full answer below

Practice focus

Application hosting, configuration overrides, middleware, authentication, dependencies, disposal, and HTTP assertions.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Why this question matters
A handler can be correct while routing, middleware order, model binding, authentication, or serialization is wrong. Hosting the application catches those composition failures.
I derive a custom WebApplicationFactory<Program>, create an HttpClient, and call the API over HTTP. The test boots the real service registration and request pipeline with test configuration. I replace external services and the database only through explicit test registrations, then assert status code, headers, and response contract rather than calling the endpoint method directly.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
The factory can remove a production registration and add a controlled fake or test container. For authentication I use a test scheme that creates known claims, while separate tests still verify real token configuration where needed. I keep one factory setup readable and give each test isolated data. This level is especially useful for validation ProblemDetails, authorization policies, middleware ordering, and JSON options.
Finish
Advanced answer
Optional deeper angle for production judgment and nuance.
I decide which dependencies remain real from the risk: SQL translation usually deserves a relational database, while a payment provider can use a protocol-level stub. Shared hosts improve speed, but mutable state must be partitioned and all resources disposed. I keep one test proving the production service-registration path so overrides cannot hide a missing registration. On failure I capture application logs, correlation ids, and safe response bodies so an integration failure is diagnosable rather than just another unexpected 500.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Calling controller methods and calling it integration testing.
Replacing every registration with a mock.
Sharing one mutable database state.
Ignoring response headers and error contracts.

Red flagsCandidateSigns the approach is fragileSignals that an interviewer may treat as risky.Open details

Details

The test bypasses routing or middleware.
Production and test configuration drift silently.
A failed request exposes no application logs.

Follow-up prompts

Quick interviewer prompts inside this drill.

Follow-up questionsCandidateFollow-up questionsLikely next questions after the main answer.Open details

Details

Treat these as the next interviewer turn. Answer one out loud, then save the gap in your practice notes.

How would you replace one dependency?
How do you test authenticated endpoints?
Would you share the factory across tests?

Continue the section

Next in Testing & Quality: Database Isolation

Step 10 of 20

Isolate database integration tests

Opening drill

Concept questionMedium~7 minTesting & Quality / 9 of 20

Interview question

How would you integration-test an ASP.NET Core API with `WebApplicationFactory`?

Integration-test ASP.NET Core with WebApplicationFactory

Uses the real ASP.NET Core pipeline while replacing only environment-specific boundaries deliberately.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

Full answer below

Practice focus

Application hosting, configuration overrides, middleware, authentication, dependencies, disposal, and HTTP assertions.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Why this question matters
A handler can be correct while routing, middleware order, model binding, authentication, or serialization is wrong. Hosting the application catches those composition failures.
I derive a custom WebApplicationFactory<Program>, create an HttpClient, and call the API over HTTP. The test boots the real service registration and request pipeline with test configuration. I replace external services and the database only through explicit test registrations, then assert status code, headers, and response contract rather than calling the endpoint method directly.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
The factory can remove a production registration and add a controlled fake or test container. For authentication I use a test scheme that creates known claims, while separate tests still verify real token configuration where needed. I keep one factory setup readable and give each test isolated data. This level is especially useful for validation ProblemDetails, authorization policies, middleware ordering, and JSON options.
Finish
Advanced answer
Optional deeper angle for production judgment and nuance.
I decide which dependencies remain real from the risk: SQL translation usually deserves a relational database, while a payment provider can use a protocol-level stub. Shared hosts improve speed, but mutable state must be partitioned and all resources disposed. I keep one test proving the production service-registration path so overrides cannot hide a missing registration. On failure I capture application logs, correlation ids, and safe response bodies so an integration failure is diagnosable rather than just another unexpected 500.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Calling controller methods and calling it integration testing.
Replacing every registration with a mock.
Sharing one mutable database state.
Ignoring response headers and error contracts.

Red flagsCandidateSigns the approach is fragileSignals that an interviewer may treat as risky.Open details

Details

The test bypasses routing or middleware.
Production and test configuration drift silently.
A failed request exposes no application logs.

Follow-up prompts

Quick interviewer prompts inside this drill.

Follow-up questionsCandidateFollow-up questionsLikely next questions after the main answer.Open details

Details

Treat these as the next interviewer turn. Answer one out loud, then save the gap in your practice notes.

How would you replace one dependency?
How do you test authenticated endpoints?
Would you share the factory across tests?

Continue the section

Next in Testing & Quality: Database Isolation

Step 10 of 20

Isolate database integration tests

How would you integration-test an ASP.NET Core API with `WebApplicationFactory`?

Practice focus

Answer path

Why this question matters

Review signals

Follow-up prompts

Continue the section

How would you integration-test an ASP.NET Core API with `WebApplicationFactory`?

Practice focus

Answer path

Why this question matters

Review signals

Follow-up prompts

Continue the section