Your DevOps team has maintained a large number of EC2 instances in several AWS accounts. There are some instances that were launched for testing purposes and are no longer used. To save cost, you need to work out an approach to quickly identify the EC2 instances that have a low utilization rate, such as daily CPU utilization is 10% or less for several days. Which method is the best that you should choose? 

Answer options:

A.In CloudWatch Metrics, for each instance, create an alarm and trigger an SNS notification when CPU utilization is below 10%.
B.Create a Lambda function that checks the CPU utilization for each instance and triggers an SNS notification if the average CPU utilization rate is low.
C.In CloudWatch Logs, configure a Filter to check the usage rate of EC2 instances. Create an alarm if the utilization is low.
D.In Trusted Advisor, check the status of Low Utilization Amazon EC2 Instances which is part of Cost Optimization Checks.

Correct Answer – D
The question asks for the method that can quickly identify the EC2 instances with low utilization. Therefore it should be avoided to perform configurations or checks for each instance repeatedly.
Option A is incorrect: Because it is not the best approach if compared with Option D as it has to configure every instance.
Option B is incorrect: Because the user has to create a new Lambda function, which is not the quickest way.
Option C is incorrect: Because the CPU usage data is already in CloudWatch Metrics. So creating Filter in CloudWatch Logs is not essential. Besides, each EC2 instance needs to install a CloudWatch Logs agent.
Option D is CORRECT: Because Trusted Advisor can efficiently identify those EC2 instances with low utilization, which is the best way to be chosen. Check the below for its usage:
 https://aws.amazon.com/premiumsupport/technology/trusted-advisor/best-practice-checklist/