Question 29
Domain 3: Deployment and Orchestration of ML WorkflowsA company needs to create a central catalog for all the company's ML models. The models are in AWS accounts where the company developed the models initially. The models are hosted in Amazon Elastic Container Registry (Amazon ECR) repositories. Which solution will meet these requirements?
Correct answer: C
Explanation
Amazon SageMaker Model Registry is designed to be a central catalog for ML models, and model groups organize related models across accounts. Because the models are in Amazon ECR in the original AWS accounts, a new account can serve as the central registry while cross-account resource policies on each model group allow access from the central account.
Why each option is right or wrong
A. Configure ECR cross-account replication for each existing ECR repository. Ensure that each model is visible in each AWS account.
ECR replication copies container images, but it does not create a centralized ML model catalog.
B. Create a new AWS account with a new ECR repository as the central catalog. Configure ECR cross-account replication between the initial ECR repositories and the central catalog.
A central ECR repository stores images centrally, but lacks model registry metadata and governance capabilities.
C. Use the Amazon SageMaker Model Registry to create a model group for models hosted in Amazon ECR. Create a new AWS account. In the new account, use the SageMaker Model Registry as the central catalog. Attach a cross-account resource policy to each model group in the initial AWS accounts.
Amazon SageMaker Model Registry is the service that provides a centralized catalog for model versions, and a model group is the correct container for related models; the registry supports cross-account sharing through resource-based policies on model groups. In this scenario, the models remain in the original AWS accounts and are stored in Amazon ECR, so the central account can only discover and manage them if each source account grants access via a cross-account resource policy attached to the model group.
D. Use an AWS Glue Data Catalog to store the models. Run an AWS Glue crawler to migrate the models from the ECR repositories to the Data Catalog. Configure cross-account access to the Data Catalog.
AWS Glue Data Catalog catalogs data assets and schemas, not ML models stored as ECR container images.