A company deploys a stateful application on Amazon EC2 On-Demand Instances in multiple Availability Zones...

The correct answer isBbecause the company has apredictable baseline workloadthat will run forat least 1 year, plus atemporary 2-week increasein demand during a holiday period. The most cost-effective approach is to use a long-term discount option for the steady-state workload and a flexible pricing model for the short-term burst capacity.

AnEC2 Instance Savings Planis a strong fit for the existing workload because it provides cost savings for committed EC2 usage while matching a stable, predictable application footprint. Since the company is already using EC2 and the workload pattern is known, this is an efficient commitment for the baseline capacity. For the additional temporary holiday demand,On-Demand Instancesare appropriate because the extra capacity is needed only for a short duration. This avoids overcommitting to capacity that will not be used after the holiday period ends.

Option A is less cost-effective because it misses the savings opportunity for the predictable baseline workload. Option C is incorrect because the application is described asstateful, which makesSpot Instancesa poor fit due to potential interruption. Option D is incorrect because purchasing a 12-month commitment for both the normal workload and the temporary holiday increase would likely result in paying for excess committed usage after the 2-week peak ends.

AWS cost optimization guidance recommends usingdiscounted commitment models for steady-state usageandOn-Demandfor short-lived, temporary spikes. Therefore, the best answer is to use anEC2 Instance Savings Planfor the baseline andOn-Demand Instancesfor the seasonal increase.