Databricks Machine Learning Associate Question 427
Single answerYou are working on a machine learning project in Databricks and want to create a feature store table that is available across workspaces using Unity Catalog. What is a necessary step to ensure the feature store table is created successfully in Unity Catalog?
- A
Define the table schema and specify the catalog and schema during the creation process.
- B
Store the feature table in a local file system before uploading it to Unity Catalog.
- C
Enable Unity Catalog for your Databricks workspace and assign proper permissions to the target schema.
- D
Use the Delta format to store the feature data but skip specifying the catalog name.
Show answer and explanation
Correct answer: C
Explanation
Unity Catalog is a centralized governance layer for managing data and metadata in Databricks. To create a feature store table in Unity Catalog, you must first enable Unity Catalog for your workspace and assign appropriate permissions to the target schema. Without this setup, you cannot create or manage feature store tables in Unity Catalog.
- A. Incorrect.
Defining the schema and specifying the catalog and schema is necessary, but this alone will not ensure the feature store table is created in Unity Catalog. Unity Catalog must be enabled and permissions configured.
- B. Incorrect.
Unity Catalog does not support storing feature tables directly from a local file system. Tables must be created directly in a catalog managed by Unity Catalog.
- C. Correct.
Enabling Unity Catalog and assigning proper permissions to the target schema is a necessary step to use Unity Catalog for feature store tables. Without this, you cannot create or manage tables in Unity Catalog.
- D. Incorrect.
Using the Delta format is required for feature store tables, but skipping the catalog name will result in the table being created in the default metastore rather than Unity Catalog.