How Do You Get Notified When EC2 CPU Utilization Exceeds 80%?

Why the Answer Is Correct

Option A creates a custom Amazon CloudWatch alarm on the EC2 CPUUtilization metric with a threshold of 80% and configures the alarm to publish a notification to an Amazon SNS topic. CPUUtilization is a standard, out-of-the-box EC2 metric that CloudWatch collects automatically, so no agents or custom code are required. When utilization crosses 80%, the alarm transitions to the ALARM state and SNS fans the notification out to subscribers via email, SMS, or mobile push. This is the fully managed, near-real-time, serverless alerting pattern that AWS explicitly recommends for metric-based monitoring.

Why the Other Options Are Wrong

Option B fails because AWS CloudTrail is an audit service that records API calls and account activity; it has no concept of performance metrics and cannot alarm on CPU utilization. Option C uses a cron job that polls every 15 minutes, which delays detection by up to 15 minutes, adds operational overhead on the instance itself, and violates the AWS best practice of using managed services instead of custom scripts. Option D is doubly wrong: CloudTrail logs do not contain a CPUUtilization metric (that data exists only in CloudWatch), and a Lambda function polling every 15 minutes adds unnecessary cost, latency, and complexity compared with a native alarm.

Community Comment Notes

The vote distribution is unanimous — 100% of community votes went to A — indicating strong consensus with no dissenting interpretations. The most-liked comment explains that CloudWatch is the AWS service purpose-built for monitoring metrics like CPU utilization and that an 80% threshold alarm publishing to SNS is the most straightforward and efficient method. The remaining comments are short confirmations that A is correct, reinforcing that this is a straightforward service-selection question rather than a nuanced architecture trade-off.