A company is using an AWS Lambda function to monitor the metrics from an ML model. An ML engineer needs to implement a solution to send an email message when the metrics breach a threshold. Which…

Question

Options

• ALog the metrics from the Lambda function to AWS CloudTrail. Configure a CloudTrail trail to send
• BLog the metrics from the Lambda function to Amazon CloudFront. Configure an Amazon
• CLog the metrics from the Lambda function to Amazon CloudWatch. Configure a CloudWatch
• DLog the metrics from the Lambda function to Amazon CloudWatch. Configure an Amazon

• A
  6% (2)
• B
  6% (2)
• C
  86% (31)
• D
  3% (1)

Explanation

Option C is correct because the standard AWS pattern for threshold-based alerting is: Lambda → CloudWatch Metrics → CloudWatch Alarm → SNS → Email. CloudWatch is purpose-built for collecting and monitoring operational metrics, and CloudWatch Alarms natively trigger Amazon SNS notifications (which deliver email) when a metric crosses a defined threshold - this is the intended, fully managed solution.

Why the distractors are wrong:

• A (CloudTrail): CloudTrail is an audit log for AWS API calls and governance - it records who did what in your account, not custom application metrics, so it cannot monitor ML model output values.
• B (CloudFront): CloudFront is a Content Delivery Network for caching and serving web content globally - it has no role in metrics collection or alerting pipelines.
• D: Though similar to C (also uses CloudWatch for logging), D likely proposes a less direct path - such as using EventBridge or skipping CloudWatch Alarms - making it a less complete or incorrect implementation of the alerting pattern.

Memory tip: Think "CloudWatch → Alarm → SNS = Watch, Warn, Notify." Whenever an AWS scenario involves monitoring a metric and triggering a notification, the answer almost always chains these three services together.

No community discussion yet for this question.