A failed Glue job leaves a dashboard showing yesterday's data.
Notify immediately and stop publication of stale output.
Which monitoring approach is most appropriate?
Match Glue Job State Change FAILED events in EventBridge and connect them to notification and publication-stop paths.
Detailed explanation
This does not detect the failure or prevent an old dataset from being served.
This does not detect the failure or prevent an old dataset from being served.
A failure event can trigger SNS notification and a workflow failure branch before stale data is published.
A failure event can trigger SNS notification and a workflow failure branch before stale data is published.
A start event says nothing about the eventual job result.
A start event says nothing about the eventual job result.
Manual review delays detection and recovery.
Manual review delays detection and recovery.
Check official documentation
Compare the following points with the official documentation.
AWS Glue job state change events、Amazon EventBridge、SNSの公式説明を確認する。
What to confirm
開始イベントと完了失敗イベントを区別できる。
Key points
- Glue Job State
- EventBridge
- SNS
- Publication gate
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
Freshness signaling
Expose the last successful dataset time so stale data is not silently presented as current.
Recovery evidence
Include execution ID and log link so a corrected run can be tracked.