You need an automated way to stop any EC2 instance that averages under 10% CPU utilization for a continuous 60-minute period. Which approach is the MOST operationally efficient?

The most operationally efficient approach is to use Amazon CloudWatch alarms. CloudWatch is designed for monitoring and can directly trigger actions based on metric thresholds. Creating an alarm per instance with a 1-hour period and a 10% CPU utilization threshold, then attaching an EC2 stop action, automates the entire process without requiring custom code or agents for this specific task. The other options are less efficient: Deploying a cron task on every EC2 instance introduces management overhead for script deployment, maintenance, and security on each instance. Installing the CloudWatch unified agent is unnecessary here as basic EC2 CPU utilization metrics are already available via CloudWatch without an agent. The agent is for custom metrics or detailed OS metrics. Using AWS Systems Manager Run Command for polling and stopping would require custom scripting and scheduling, which is more complex and less integrated than a native CloudWatch alarm action.