Snowflake Certification Practice Test

Which of the following is not a characteristic of "automatic micro-partitions" in Snowflake?

1. New partitions created on logical properties

2. Avoids skews between partitions

3. Automatically complete on the ingestion order of the data

4. Maintains correlations between columns for pruning

The correct answer is: New partitions created on logical properties

The characteristic that is not associated with automatic micro-partitions in Snowflake is the notion that new partitions are created solely based on logical properties. In reality, Snowflake's automatic micro-partitioning system creates new micro-partitions based on the actual data ingestion patterns rather than purely logical properties. This means that micro-partitions are formed based on the distribution and structure of incoming data, which optimizes the storage and retrieval process, rather than just theoretical or logical constructs. In contrast, the other options reflect true characteristics of how automatic micro-partitioning operates. The system helps to avoid skews between partitions by evenly distributing data across the partitions based on their size and content. Additionally, it maintains an ordered ingestion process, which can influence how micro-partitions are filled and subsequently utilized for queries. Finally, Snowflake employs internal mechanisms to maintain correlations between columns, which enhances performance during query optimization and pruning. Thus, the focus on logical properties as the sole basis for partitioning is a misunderstanding of how Snowflake's micro-partitioning effectively functions.