You need a deployment stage for a serverless app (AWS Lambda) that minimizes customer impact from failed deployments and provides monitoring for issues. Which deployment configuration meets these goals?

The correct option leverages AWS CodeDeploy's built-in canary deployment strategies for Lambda, specifically Canary10Percent15Minutes, which shifts 10% of traffic to the new version for 15 minutes. This allows for real-time monitoring with CloudWatch alarms before a full rollout, minimizing customer impact from potential issues. AWS SAM simplifies defining serverless applications and integrates well with CodeDeploy. The other options are less effective. Using CloudFormation with manual approvals introduces delays and doesn't automate progressive traffic shifting. While CloudFormation with RoutingConfig on an alias can manage traffic, it requires more manual configuration for canary-style deployments compared to CodeDeploy. CodeBuild for testing and updating aliases is a valid step, but it lacks the automated, managed traffic shifting and rollback capabilities that CodeDeploy provides out-of-the-box for minimizing impact during a deployment.