Explanation:
The question requires deploying a model with two key requirements: (1) monitoring costs at the endpoint and deployment level, and (2) Microsoft Entra ID authentication. Managed online endpoints (Option D) provide built-in cost monitoring capabilities at both endpoint and deployment levels, which AKS deployments (Options A and B) do not offer natively. For authentication, Microsoft Entra ID integration requires token-based authentication, which is configured using the token_auth_mode parameter set to true in managed online endpoints. While Option C also uses managed online endpoints, it incorrectly suggests using the auth_mode parameter instead of token_auth_mode for Entra ID authentication. The community discussion confirms that token authentication is required for Entra ID and that managed endpoints support the necessary cost monitoring features.
Ultimate access to all questions.
You have an Azure Machine Learning workspace and need to deploy a trained model as an online endpoint. You must monitor costs at the endpoint and deployment level, and users are required to authenticate using Microsoft Entra ID.
What should you do?
A
Deploy the model to Azure Kubernetes Service (AKS). During deployment, set the token_auth_mode parameter of the target configuration object to true.
B
Deploy the model to Azure Kubernetes Service (AKS). During deployment, set the auth_mode parameter to configure the authentication type.
C
Deploy the model to a managed online endpoint. During deployment, set the auth_mode parameter to configure the authentication type.
D
Deploy the model to a managed online endpoint. During deployment, set the token_auth_mode parameter of the target configuration object to true.
No comments yet.