A compute cluster consists of 20 EC2 instances: 2 always-on control nodes and 18 task nodes used for processing. Control nodes must run 24/7; task nodes run about 4 hours per day and can be started by the control nodes. All instances are currently On-Demand. How can you reduce cost while meeting availability and runtime requirements? (Choose two.)

EC2 Instance Savings Plans offer significant discounts (up to 72%) on EC2 usage in exchange for a commitment to a consistent amount of compute usage (e.g., $10/hour) for a 1- or 3-year term. This is ideal for the two always-on control nodes, as their 24/7 operation guarantees consistent usage, maximizing cost savings without affecting availability. Spot Instances are a cost-effective option for the task nodes because they can tolerate interruptions. Spot Instances can offer discounts of up to 90% compared to On-Demand prices. Since the task nodes only run 4 hours a day and can be started by control nodes, they are well-suited for Spot Instances, with a fallback to On-Demand ensuring availability if Spot capacity is interrupted. Running control nodes on Dedicated Hosts is unnecessary and expensive, as Dedicated Hosts are primarily for specific licensing requirements. Purchasing Reserved Instances for the task nodes is less flexible and potentially less cost-effective than Spot Instances, given their intermittent usage. Running control nodes as Spot Instances is risky because their 24/7 availability is critical, and Spot Instances can be interrupted with short notice.