writing-unit-tests

To write unit tests for SCT without external services, isolate your code by mocking AWS, GCE, and network calls using unittest.mock, moto, and monkeypatch. This ensures tests verify logic independently without hitting live infrastructure.

The best way to mock AWS and GCE calls in pytest is by using moto for AWS services and unittest.mock for GCE. These tools allow you to patch external interactions, ensuring your unit tests remain isolated and repeatable.

Debug slow or flaky pytest runs by leveraging targeted debugging strategies and pytest fixtures. Isolate external dependencies with mocking to prevent network timeouts and identify non-deterministic behaviors causing test instability.

Yes, you can use monkeypatch and unittest.mock together in SCT testing. Monkeypatch handles safe attribute patching during test execution, while unittest.mock provides robust mock objects for verifying complex interactions.

Available pytest fixtures for SCT unit tests provide setup and teardown mechanisms for your test environment. Consult the test infrastructure reference to understand specific markers and fixtures designed to isolate test logic.

SCT unit tests fail without mocked network calls because they attempt to reach live external infrastructure. Isolating tests using monkeypatch or unittest.mock prevents these failures by ensuring no real network requests are executed.