QuestionQ64

VerifiedDesign High-Performing Architectures

A company is moving a legacy application to AWS. The application runs a batch job once every hour and is CPU-intensive. On an on-premises server, the batch job takes 15 minutes on average. That server has 64 virtual CPUs (vCPUs) and 512 GiB of memory.

Which solution will execute the batch job within 15 minutes with the LEAST operational overhead?

Explanation

AWS Batch is the AWS service designed specifically for batch and compute-intensive jobs. It is a fully managed batch computing service that plans, schedules, and runs containerized batch workloads, and it can scale compute capacity for those jobs automatically. The Amazon EC2 compute option in AWS Batch can provide the larger CPU and memory capacity that a workload comparable to a 64-vCPU, 512-GiB server may require. Lambda does not provide that class of per-job compute capacity, and ECS with Fargate is a general container runtime rather than a purpose-built batch scheduler. Therefore, AWS Batch on Amazon EC2 best satisfies the performance target with the least operational overhead among the listed choices.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!