A company has a production Amazon EKS cluster with managed node groups that use On-Demand Instances. The company needs a separate EKS cluster for development testing that will be used infrequently to test application resiliency. The EKS cluster must manage all nodes. Which solution meets these requirements MOST cost-effectively?

Creating a managed node group with only Spot Instances is the most cost-effective solution for an infrequently used development testing cluster. Spot Instances offer significant cost savings (up to 90% compared to On-Demand) and are suitable for fault-tolerant workloads like development testing where interruptions are acceptable. EKS managed node groups simplify node management, automatically handling updates, patching, and scaling. Using two managed node groups (On-Demand and Spot) would be more expensive than necessary for an infrequently used, fault-tolerant development cluster. An Auto Scaling group with a launch configuration is a more manual approach to node management compared to EKS managed node groups, increasing operational overhead. A managed node group with only On-Demand Instances would be significantly more expensive than using Spot Instances, negating the cost-effectiveness requirement.