Quality Control & Assurance Flashcards

A developer needs to test how an Azure Service Bus consumer handles poison messages. What is the best practice in a test environment?

Answer: Send malformed messages and assert the consumer moves them to the dead-letter sub-queue after max delivery attempts

Intentionally sending invalid messages and verifying they land in the dead-letter queue validates the poison-message handling path without impacting production queues.

Which Azure Cosmos DB feature helps you validate that changes to your data model do not break existing queries before deploying to production?

Answer: Automatic indexing policy validation via the Azure portal query explorer

The Azure portal's Data Explorer lets you run queries against a copy of your container with a new indexing policy to verify correctness before applying changes.

When writing integration tests for an Azure SQL Database, which approach best isolates each test from side effects of other tests?

Answer: Wrap each test in a transaction and roll it back in teardown

Wrapping each test in a transaction that is rolled back on teardown ensures the database returns to a clean state without permanently modifying data.

You need to enforce that no NuGet package with a known critical vulnerability is used in your Azure DevOps build. Which component enables this?

Answer: A pipeline task running 'dotnet list package --vulnerable' and failing on critical findings

'dotnet list package --vulnerable' queries NuGet security advisories and the pipeline task can be scripted to exit with a non-zero code if critical vulnerabilities are found.

In Azure Application Insights, what does 'Sampling' do and why is it relevant to QA?

Answer: Sampling reduces the volume of telemetry ingested by keeping a representative subset, which can mask infrequent errors in QA analysis

Adaptive or fixed-rate sampling discards a percentage of telemetry, so rare error events may not appear in reports, which is important to understand when analyzing QA results.

Which HTTP status code returned by an Azure API Management policy or backend signals that a request was throttled, and how should an automated test assert this?

Answer: 429 Too Many Requests — assert the response status code equals 429

HTTP 429 is the standard throttling response; tests should assert on the exact status code to confirm rate-limit policies are enforced correctly.

A QA engineer wants to validate that an Azure Event Grid subscription delivers events to a webhook endpoint within an SLA of 5 seconds. What is the most practical testing strategy?

Answer: Publish a test event and measure the elapsed time until the webhook receives the callback, failing if it exceeds 5 seconds

Publishing a synthetic event from a test harness and timestamping webhook receipt provides an end-to-end latency measurement that can be asserted against the SLA threshold.