Answer: A column that is most actively used in selective filters, A column that is most actively used in join predicates

Snowflake recommends using columns that are most actively used in selective filters (D) and columns that are most actively used in join predicates (E) as clustering keys. This is because clustering keys optimize query performance by organizing data into micro-partitions based on these frequently filtered or joined columns, allowing Snowflake to efficiently prune irrelevant micro-partitions during query execution. The community discussion shows 100% consensus on DE, with references to Snowflake's official documentation on clustering keys. Options A (VARIANT column), B (very low cardinality), and C (very high cardinality) are not recommended - VARIANT columns contain semi-structured data that is inefficient for clustering, very low cardinality columns provide minimal pruning benefit, and very high cardinality columns can lead to excessive micro-partitions without significant performance gains.

Author: LeetQuiz Editorial Team