MLA-C01 Question 309
Select 3You are tasked with deploying a machine learning model as a scalable API endpoint. The model is packaged as a container image. Which AWS services can you use to host the containerized model and provide scalability and high availability?
- A
Amazon ECS with AWS Fargate
- B
Amazon SageMaker
- C
AWS Lambda
- D
AWS Elastic Beanstalk
- E
Amazon RDS
Show answer and explanation
Correct answers: A, B, D
Explanation
Amazon ECS with AWS Fargate, Amazon SageMaker, and AWS Elastic Beanstalk are valid services for hosting containerized machine learning models with scalability and high availability. Amazon ECS with Fargate allows you to run containers without managing servers, Amazon SageMaker provides dedicated support for ML model deployment, and Elastic Beanstalk simplifies application deployment for containerized workloads. AWS Lambda, while supporting containers, is not ideal for long-running APIs, and Amazon RDS is unrelated to hosting containers as it is a database service.
- A. Correct.
Amazon ECS with AWS Fargate is a suitable service for running containerized workloads. It provides scalability and high availability without requiring you to manage the underlying EC2 instances.
- B. Correct.
Amazon SageMaker is specifically designed for machine learning workloads. It supports deploying models as endpoints by using containers and offers built-in scalability options.
- C. Incorrect.
AWS Lambda is not an ideal choice for hosting containerized APIs that require long-running processes or large models. While it supports container images, it is better suited for short-lived, event-driven computing.
- D. Correct.
AWS Elastic Beanstalk supports containerized applications and can automatically handle deployment, scalability, and monitoring, making it a valid option for hosting containerized APIs.
- E. Incorrect.
Amazon RDS is a managed database service and not suitable for hosting containerized machine learning models or APIs.