A workload runs on Amazon EC2 Spot Instances in an Auto Scaling group. The launch template currently specifies two placement groups and a single instance type. Recently, Auto Scaling reported launch failures, causing longer wait times for users. To improve reliability of the Spot-based workload, which change should you make?

The correct option improves reliability by diversifying instance types, making it easier for the Auto Scaling group to find available Spot capacity. Attribute-based instance type selection allows you to define instance requirements (e.g., vCPU, memory) rather than specific instance types, giving the Auto Scaling group more flexibility. Using a new launch template version ensures a controlled update without disrupting existing configurations. Replacing the launch template with a launch configuration is incorrect because launch configurations are an older feature and do not support attribute-based instance type selection. Increasing placement groups alone doesn't address the core issue of limited instance type availability. Using a single, larger instance type would further restrict options and potentially worsen launch failures.