Exam AWS Certified Solutions Architect - Associate SAA-C03 topic 1 question 483 discussion

Lambda supports only Linux-based container images. https://docs.aws.amazon.com/lambda/latest/dg/images-create.html

By using Amazon ECS on AWS Fargate, you can run the job in a containerized environment while benefiting from the serverless nature of Fargate, where you only pay for the resources used during the job's execution. Creating a scheduled task based on the container image of the job ensures that it runs every 10 minutes, meeting the required schedule. This solution provides flexibility, scalability, and cost-effectiveness.

Lambda can be in .net container, and it is the cheaper than any other options.

For B: Even though AWS Batch is free, I think it's more suitable for large-scale or relatively complex workloads, not lightweight periodic jobs. So B is not cost-effective (The cost depends on the compute resources you provision to execute the jobs, and maybe "cost-effective" somewhat involves less operational overhead because you have spent your money) If you require queueing for task prioritization or processing a backlog of tasks, or work on data processing over datasets where Batch excels, AWS Batch would be better!

A A A A A A

selected answer : A AWS Lambda now supports .NET 6 as both a managed runtime and a container base image

https://aws.amazon.com/about-aws/whats-new/2022/02/aws-lambda-adds-support-net6/

The question is weirdly phrased for .Net based containers. "A company containerized a Windows job that runs on .NET 6 Framework under a Windows container." This could mean that the job requires .Net 6 Framework OR it could mean the job requires Windows and .Net Framework 6. If the job is just based on .Net 6 then Lambda can run it. I am just a bit cautious about language because other parameters fall under Lambda. Question may have been wrongly quoted here.

I guess this is an old question from before August 2023, when AWS Batch did not support Windows containers, while ECS already did since September 2021. Thus it would be C, though now B does also work. Since both Batch and ECS are free, we'd pay only for the Fargate resources (which are identical in both cases), now B and C would be correct. A doesn't work because Lambda still does not support Windows containeres. D doesn't make sense because the container would have to run 24/7

I think that Batch with Fargate is more cheaper than ECS.

Batch supports fargate now

aws batch supports fargate

C works. For A, the lambda support container image, but the container image much implement the Lambda Runtime API.

As they support Batch on Fargate now (Aug 2023), the correct answer should be B?

https://docs.aws.amazon.com/lambda/latest/dg/csharp-image.html#csharp-image-clients

C is the most cost-effective solution for running a short-lived Windows container job on a schedule. Using Amazon ECS scheduled tasks on Fargate eliminates the need to provision EC2 resources. You pay only for the duration the task runs. Scheduled tasks handle scheduling the jobs and scaling resources automatically. This is lower cost than managing your own scaling via Lambda or Batch. ECS also supports Windows containers natively unlike Lambda (option A). Option D still requires provisioning and paying for full time EC2 resources to run a task scheduler even when tasks are not running.

August 2023, AWS Batch now support Windows container https://docs.aws.amazon.com/batch/latest/userguide/fargate.html#when-to-use-fargate