The data workload is constant and predictable.

C and D are impossible because you don't set or specify RCUs and WCUs in on-demand mode. A is wrong because there is no indication of "infrequent access", and "the data workload is constant", there is no different between the current and the "forecasted" workload.

I think it is not possible to set Read Capacity Units(RCU)/Write Capacity Units(WCU) in on-demand mode.

predictable/constant => provisioned mode. On-demand mode is more suitable for workloads that are unpredictable and can vary widely from minute to minute. The use case is not for Standard-IA which is described here: https://aws.amazon.com/dynamodb/standard-ia/ => Option B

I rule out A because of this 'Standard-Infrequent Access ', clearly the company uses applications to analyze the data. The data workload is constant and predictable making provisioned mode the best option.

Option B lacks the cost benefits of Standard-IA. Option C uses more expensive on-demand pricing. Option D does not actually allow reserving capacity with on-demand mode. So option A leverages provisioned mode, Standard-IA, and reserved capacity to meet the requirements in a cost-optimal way.

A is correct!

예측가능..

Option C is the most cost-effective solution for this scenario. In on-demand mode, DynamoDB automatically scales up or down based on the current workload, so the company only pays for the capacity it uses. By setting the RCUs and WCUs high enough to accommodate changes in the workload, the company can ensure that it always has the necessary capacity without overprovisioning and incurring unnecessary costs. Since the workload is constant and predictable, using provisioned mode with reserved capacity (Options A and D) may result in paying for unused capacity during periods of low demand. Option B, using provisioned mode without reserved capacity, may result in throttling during periods of high demand if the provisioned capacity is not sufficient to handle the workload.

"The data workload is constant and predictable." https://docs.aws.amazon.com/wellarchitected/latest/serverless-applications-lens/capacity.html "With provisioned capacity you pay for the provision of read and write capacity units for your DynamoDB tables. Whereas with DynamoDB on-demand you pay per request for the data reads and writes that your application performs on your tables."

The data workload is constant and predictable, then, isn't on-demand mode. DynamoDB Standard-IA is not necessary in this context

The problem with (A) is: “Standard-Infrequent Access“. In the question, they say the company has to analyze the Data. That’s why the Correct answer is (B)

workload is constant

As the numbers are already known