Which statement about materialized view limitations is true?

• A. Joins are not supported; materialized views query only a single table
• B. They can join multiple tables
• C. They can include window functions
• D. They can query across multiple databases

Answer: (D)

Materialized views are precomputed results built from a defining query, and the way they access data is what this item is testing. The true statement is that a materialized view can query across multiple databases. In Snowflake you can reference objects from different databases in the MV’s defining query by using fully qualified names (for example, database.schema.table), as long as you have the necessary privileges. This enables cross-database analytics with the performance benefits of a precomputed result.

The other statements misstate common capabilities: materialized views can involve joins across multiple tables rather than being limited to a single table, so the claim that joins aren’t supported isn’t accurate. The specifics about window functions aren’t the tested limitation here, and the key point this item highlights is the cross-database querying capability.