A company runs an image-processing application on a serverless infrastructure. Each processing job runs in a single AWS Lambda execution. A sysops administrator is tasked with ensuring there is enough capacity to run 500 simultaneous jobs even if other Lambda functions are being run for other applications. The administrator has already increased service limits within the Region.
Which action should be taken?

A. Configure a dead-letter queue to retry any throttled executions
B. Modify the memory settings on the Lambda function to allow for 500 parallel executions
C. Move the image-processing logic to AWS Step Functions
D. Set the reserved concurrency for the image-processing Lambda function to 500