A company wants to use AWS to scale up the number of its long-running critical...

Use AWS Lambda functions to perform simulations. Store structured data in Amazon RDS. Store unstructured data in Amazon S3.

Use AWS Batch to perform simulations. Store structured data in Amazon RDS. Store unstructured data in Amazon S3.

Use Amazon EC2 Spot Instances to perform simulations. Store structured data in Amazon RDS. Store unstructured data in Amazon DynamoDB.

Use AWS Batch to perform simulations. Store all data in an Amazon EBS volume.

B

AWS Batch is the correct service for large-scale parallel simulation jobs that run for a long time. It can manage batch compute environments, job queues, and scheduling without requiring the company to manually coordinate simulation workers. Because the simulations cannot be stopped, the compute environment should use reliable capacity rather than Spot-only capacity. Amazon RDS is appropriate for structured simulation results because it provides managed relational storage, backups, durability, and high availability options. Amazon S3 is the correct choice for unstructured image output because it is highly durable, fault-tolerant object storage and can store files such as images efficiently. Lambda is invalid because simulations run for days, far beyond Lambda’s maximum runtime. EBS alone is not a durable, fault-tolerant shared output architecture.