A Glue job fails at night. How does the team find out quickly?
Why B is the answer
A CloudWatch alarm or EventBridge rule that sends an SNS notification
Automated alerts shorten detection time.
In detail
- An alarm or rule with SNS tells the team quickly.
- Send alerts to a group.
Use case
An illustrative scenario, not a real incident.
A Glue job fails at 2am. An EventBridge rule sends an SNS message, so the team knows before the reports run.
See it in different ways
flowchart LR A["Glue or Lambda failure"] --> B["CloudWatch Logs"] B --> C["Metric filter"] C --> D["CloudWatch alarm"] D --> E["Amazon SNS notification"]
sequenceDiagram participant S as S3 participant E as EventBridge participant F as Step Functions participant N as SNS S->>E: Object created E->>F: Start the workflow F->>F: Glue job, with retries F-->>N: Alert on failure
sequenceDiagram participant P0 as Glue job participant P1 as EventBridge participant P2 as Amazon SNS participant P3 as Team P0->>P1: Job failed P1->>P2: Publish P2->>P3: Notify
Key point: Automate failure alerts.
Why the other options are wrong
- A. The users report missing dashboards
- Users should not be your monitor.
- C. A person checks the console every morning
- That finds failures hours later.
- D. An analyst checks the job list each morning
- That finds failures hours later.