Google Professional Data Engineer Question 187
Single answerGoogle Cloud PlatformYou are working on a data analytics project where the query performance is critical for your BigQuery dataset. You are considering using materialized views to optimize performance. Which of the following statements about BigQuery materialized views is correct?
- A
Materialized views automatically refresh when the base table is updated to ensure query results are always up-to-date.
- B
Materialized views store precomputed results and can significantly reduce query latency for frequently accessed data.
- C
Materialized views support all SQL queries, including those with non-deterministic functions like CURRENT_TIMESTAMP.
- D
Materialized views are suitable for use cases where the underlying data changes frequently and low-latency updates are required.
Show answer and explanation
Correct answer: B
Explanation
Materialized views in BigQuery are designed to optimize performance by storing precomputed query results, which makes them advantageous for repetitive and complex queries. However, they do not refresh in real-time, have limitations on the type of SQL queries they support, and are not suitable for use cases requiring low-latency updates to frequently changing data.
- A. Incorrect.
This is incorrect. Materialized views do not automatically refresh in real-time when the base table is updated. Instead, they follow a scheduled refresh mechanism or require a manual refresh.
- B. Correct.
This is correct. Materialized views precompute and store query results, which helps reduce query latency and improve performance for repetitive or complex queries.
- C. Incorrect.
This is incorrect. Materialized views do not support all SQL queries. They have certain limitations, such as not supporting non-deterministic functions like CURRENT_TIMESTAMP or queries involving user-defined functions.
- D. Incorrect.
This is incorrect. Materialized views are not ideal for frequently changing data with low-latency update requirements, as their refresh mechanism is not real-time.