A Glue ETL job receives rows with invalid date formats.
Continue processing valid rows while retaining failed rows with diagnostics for later replay.
Which configuration is most appropriate?
Isolate failed rows in a dedicated S3 prefix and record the input ID, error details, and processing version.
Detailed explanation
Discarding removes the evidence needed to investigate and recover missing data.
Discarding removes the evidence needed to investigate and recover missing data.
Separating failed data preserves evidence and allows corrected rows to be replayed without stopping valid processing.
Separating failed data preserves evidence and allows corrected rows to be replayed without stopping valid processing.
A value-format error is different from authorization failure and cannot be fixed by changing IAM.
A value-format error is different from authorization failure and cannot be fixed by changing IAM.
Object generations do not identify which rows failed or why the ETL rejected them.
Object generations do not identify which rows failed or why the ETL rejected them.
Check official documentation
Compare the following points with the official documentation.
AWS Glueのエラー処理、DynamicFrameのerrorsAsDynamicFrame、S3出力の公式説明を確認する。
What to confirm
値の変換エラーとIAM権限エラーを区別し、失敗データを再処理可能にする設計を説明できる。
Key points
- Error isolation
- Replay
- Failed row
- Glue ETL
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
Failure taxonomy
Input, schema, authorization, capacity, and service failures require different responses.
Replay contract
Original data, input identity, and transformation version make a retry auditable and safe.