A vendor supplies a unit-testing product as an Amazon Machine Image (AMI) for EC2. Configuration settings are kept in Amazon DynamoDB and test outputs are saved to Amazon S3. The product requires at least three EC2 instances to run, and the testing team wants the ability to add three more EC2 instances using Spot Instances when Spot prices meet a target. As a SysOps administrator, design a highly available deployment with minimal operational maintenance. Which approach satisfies these requirements with the least overhead?

Using a single EC2 Auto Scaling group with a launch template allows for a mixed instances policy, which is the most efficient way to manage both On-Demand and Spot Instances for the same application. The launch template provides a modern, version-controlled method for specifying instance configurations, including the AMI, instance type, and Spot Instance pricing. This approach minimizes operational overhead by consolidating management into one Auto Scaling group. Launch configurations are an older method and do not support mixed instances policies as effectively. Creating two separate Auto Scaling groups would increase operational overhead and complexity compared to a single, unified group.