AWS Certified Solutions Architect - Associate
Get started today
Ultimate access to all questions.
A company is developing ML models as microservices on AWS, with each model loading 1 GB of data from S3 into memory at startup. The system handles irregular user request patterns, ranging from inactivity to thousands of concurrent requests. Which architecture is recommended for this scalable, asynchronous API-based ML model solution?
A company is developing ML models as microservices on AWS, with each model loading 1 GB of data from S3 into memory at startup. The system handles irregular user request patterns, ranging from inactivity to thousands of concurrent requests. Which architecture is recommended for this scalable, asynchronous API-based ML model solution?
Explanation:
The correct answer is D. This option leverages Amazon SQS for request queuing and Amazon ECS for processing, enabling Auto Scaling based on queue size. This meets the requirements of handling irregular usage patterns efficiently and scaling based on demand. ECS can manage clusters and services, which is more suitable for this scenario where loading 1 GB of model data into memory at startup requires persistent instances, as opposed to the ephemeral nature of AWS Lambda.