Exam SnowPro Core topic 1 question 37 discussion

A. VARIANT “null” values are not the same as SQL NULL values: In Snowflake, a VARIANT type can store semi-structured data like JSON. A JSON null is distinct from an SQL NULL. JSON null is treated as a valid value in Snowflake, while SQL NULL represents the absence of a value. B. It is recommended to do frequent, single row DMLs: This is false. Snowflake is optimized for bulk operations, and frequent single-row DMLs (e.g., INSERT, UPDATE, DELETE) are not recommended as they can lead to inefficiencies and higher costs. C. It is recommended to validate the data before loading into the Snowflake target table: This is true. Validating data before loading ensures data quality and avoids errors or inconsistencies in the target table. D. It is recommended to use staging tables to manage MERGE statements: This is true. Staging tables are often used to prepare and validate data before applying MERGE operations, which helps ensure data integrity and efficiency. Correct Choices: A, C, D

ACD are correct

ACD are correct

ACD are correct

Variant NULL values: https://docs.snowflake.com/en/user-guide/semistructured-considerations#label-variant-null

ACD is correct answer.

The result must be deterministic for merge so source can be any reference data source only constraint it should create/modify unique rows

ACD are correct

Frequent Single rows on any MPP like Snowflake has always been a no no! Function of an OLTP DB

Frequent single row DML? Hell nah. ACD is the way to go.

ACD, Correct