Databricks Generative AI Engineer Associate Question 189
Select 4You are deploying a machine learning model using Databricks Model Serving. To ensure that only authorized users and applications can access the model serving endpoints, which of the following approaches should you implement?
- A
Configure access control lists (ACLs) for the model serving endpoint.
- B
Use API tokens to authenticate requests to the model serving endpoint.
- C
Set up IP allowlists to restrict access to specific ranges of IP addresses.
- D
Enable public access to the model serving endpoint for easier integration.
- E
Integrate the serving endpoint with an identity provider (IdP) for Single Sign-On (SSO).
Show answer and explanation
Correct answers: A, B, C, E
Explanation
To secure access to model serving endpoints in Databricks, you should implement a combination of access control mechanisms such as ACLs, API tokens, IP allowlists, and integration with identity providers for SSO. Public access should be avoided to ensure that only authorized users and applications can interact with the endpoint.
- A. Correct.
Configuring access control lists (ACLs) is an essential step to limit access to specific users or groups for the model serving endpoint.
- B. Correct.
API tokens provide a secure way to authenticate and authorize requests to the model serving endpoint, ensuring only authorized clients can access it.
- C. Correct.
IP allowlists add an additional layer of security by restricting access to the endpoint to known, trusted IP ranges.
- D. Incorrect.
Enabling public access to the endpoint would expose it to unauthorized access, which is not a secure practice for sensitive model serving applications.
- E. Correct.
Integrating with an identity provider (IdP) for SSO allows for centralized and secure user authentication, aligning with enterprise security standards.