MLA-C01 Practice Question 100: Training Failure Notifications
AWS Certified Machine Learning Engineer – Associate (MLA-C01) · Test 10 · question 10 of 10
An ML team wants an email within about a minute whenever any SageMaker AI training job in its account ends with a Failed status, whether a pipeline, a notebook, or a scheduled script started it. The email should include why the job failed. The team does not want to run any polling code.
What should the team configure?
Show answer and explanation
Answer:
SageMaker AI sends events to Amazon EventBridge when a training job's status changes. When TrainingJobStatus is Failed, the event includes a FailureReason field describing why. An EventBridge rule with an event pattern matching failed training job state changes can target an Amazon SNS topic directly, so every failed job produces an email with the reason, whatever started the job, and nothing polls (option C).
- A is wrong because CreateTrainingJob records a job being *created*, not how it ended. The metric filter would count every new job, and CloudTrail events do not carry the failure reason. - B is wrong because training metrics are per job and specific to each algorithm, so jobs would need alarms set up in advance. A metric that stops reporting also cannot tell a failed job from one that finished successfully. - D is wrong because a scheduled function that calls ListTrainingJobs is the polling code the team wants to avoid, and it adds up to five minutes of delay.
Reference: Events that Amazon SageMaker AI sends to Amazon EventBridge