Which statement about compression within micro-partitions is true?

• A. Columns are compressed individually within micro-partitions.
• B. Compression algorithm is fixed for all micro-partitions.
• C. Micro-partitions are compressed as a whole.
• D. Snowflake does not compress micro-partitions.

Answer: (A)

Compression within micro-partitions is done per column. Snowflake stores data in micro-partitions in a columnar format, and each column’s values are encoded and compressed separately using techniques suited to that column’s data pattern. This per-column approach lets the system tailor compression to the distribution of values in each column, achieving better space efficiency and faster columnar access. The idea that the compression algorithm is fixed for all micro-partitions ignores that encoding can vary by column and data, so it isn’t accurate. Compressing the micro-partition as a single unit would miss the columnar approach Snowflake uses, and saying there’s no compression isn’t true because compression is a fundamental part of how data is stored.