Order amounts can be NULL or negative.
Quarantine invalid rows before writing the analytical table, while aggregating only valid data.
Which processing design is most appropriate?
Apply Glue ETL validation rules and write invalid rows to a separate output prefix.
Detailed explanation
CloudTrail records AWS API activity and does not validate business-data values.
CloudTrail records AWS API activity and does not validate business-data values.
IAM controls permissions and does not validate record contents.
IAM controls permissions and does not validate record contents.
Validation separates normal and invalid data while preserving evidence for investigation and replay.
Validation separates normal and invalid data while preserving evidence for investigation and replay.
This can hide defects and loses values needed for investigation or audit.
This can hide defects and loses values needed for investigation or audit.
Check official documentation
Compare the following points with the official documentation.
AWS Glueの変換、Data Quality、エラー行出力の公式説明を確認する。
What to confirm
値を無条件に捨てるのではなく、検証・隔離・再処理の流れを説明できる。
Key points
- Data quality
- NULL
- Invalid value
- Error quarantine
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
Quality rules
Completeness, range, format, and uniqueness should be stated as measurable rules.
Lossless failure handling
Keeping the original row and reason avoids hiding defects and supports controlled reprocessing.