A Python transformation function is changed.
Before deployment, test representative valid, invalid, and boundary data without using the full production dataset.
Which implementation is most appropriate?
Run repeatable unit tests with fixed fixtures in CodeBuild and stop deployment when a quality check fails.
Detailed explanation
This risks data damage, exposes sensitive data, and is hard to reproduce.
This risks data damage, exposes sensitive data, and is hard to reproduce.
Small reproducible inputs verify transformation results and error handling before deployment.
Small reproducible inputs verify transformation results and error handling before deployment.
Broader permissions do not prove that transformed data is correct.
Broader permissions do not prove that transformed data is correct.
A cost metric does not validate transformation correctness.
A cost metric does not validate transformation correctness.
Check official documentation
Compare the following points with the official documentation.
AWS CodeBuild、CodePipeline、Python pytestの公式説明を確認する。
What to confirm
本番目視確認と再現可能な自動テストの違いを説明できる。
Key points
- CodeBuild
- pytest
- Fixture
- Unit test
Notes
- Environment: AWS公式DEA-C01試験ガイドと各サービスの公式ドキュメントの確認
- Confirm current AWS service behavior in the official DEA-C01 documentation.
Foundation review
Representative data
Fixtures should cover valid, invalid, null, boundary, and privacy-sensitive cases.
Fail-fast delivery
A failed test should stop promotion and preserve the artifact and result for diagnosis.