AWS Certified Solutions Architect – Professional (SAP-C02) — Question 236

A company has an application that runs on Amazon EC2 instances in an Amazon EC2 Auto Scaling group. The company uses AWS CodePipeline to deploy the application. The instances that run in the Auto Scaling group are constantly changing because of scaling events.

When the company deploys new application code versions, the company installs the AWS CodeDeploy agent on any new target EC2 instances and associates the instances with the CodeDeploy deployment group. The application is set to go live within the next 24 hours.

What should a solutions architect recommend to automate the application deployment process with the LEAST amount of operational overhead?

Answer options

• A. Configure Amazon EventBridge to invoke an AWS Lambda function when a new EC2 instance is launched into the Auto Scaling group. Code the Lambda function to associate the EC2 instances with the CodeDeploy deployment group.
• B. Write a script to suspend Amazon EC2 Auto Scaling operations before the deployment of new code. When the deployment is complete, create a new AMI and configure the Auto Scaling group's launch template to use the new AMI for new launches. Resume Amazon EC2 Auto Scaling operations.
• C. Create a new AWS CodeBuild project that creates a new AMI that contains the new code. Configure CodeBuild to update the Auto Scaling group’s launch template to the new AMI. Run an Amazon EC2 Auto Scaling instance refresh operation.
• D. Create a new AMI that has the CodeDeploy agent installed. Configure the Auto Scaling group’s launch template to use the new AMI. Associate the CodeDeploy deployment group with the Auto Scaling group instead of the EC2 instances.

Correct answer: D

Explanation

The correct answer is D because creating a new AMI with the CodeDeploy agent pre-installed simplifies the deployment process for new instances in the Auto Scaling group, allowing them to automatically be associated with the deployment group. Options A and C involve additional steps like using Lambda or CodeBuild, which introduce more complexity and potential points of failure. Option B requires manual intervention to pause and resume scaling operations, which increases operational overhead.