A Glue job sometimes fails because of temporary service throttling.
Retry transient errors with exponential backoff, but notify and stop for invalid input or other permanent errors.
Which configuration is most appropriate?
Configure Step Functions Retry and Catch by error class, with bounded attempts and a recovery branch.
Detailed explanation
Retry handles transient faults while Catch sends unrecoverable errors to a separate notification or isolation path.
Retry handles transient faults while Catch sends unrecoverable errors to a separate notification or isolation path.
Invalid input and missing permissions keep failing, increasing cost and delay without solving the cause.
Invalid input and missing permissions keep failing, increasing cost and delay without solving the cause.
More audit history does not control workflow retries.
More audit history does not control workflow retries.
Deleting the input hides the cause and removes the ability to replay it.
Deleting the input hides the cause and removes the ability to replay it.
Check official documentation
Compare the following points with the official documentation.
Step FunctionsのRetry、BackoffRate、MaxAttempts、Catchの公式説明を確認する。
What to confirm
無限再試行ではなく、エラー分類と上限付き再試行を説明できる。
Key points
- Retry
- Catch
- Exponential backoff
- Permanent error
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
Transient failure
Throttling and temporary network faults may recover after a bounded delay.
Permanent failure
Invalid data or missing permissions require correction instead of repeated attempts.