In a Snowflake query profile, what does the metric Partitions Scanned versus Partitions Total indicate?

• A. The partitions read by the query vs the total partitions in the table
• B. The number of partitions pruned by the optimizer
• C. The number of warehouses involved in the query
• D. The size of each partition scanned

Answer: (A)

This metric is about data pruning and how much of the table’s partition space the query touches. Snowflake stores data in partitions, and whether a query reads a partition depends on the filters and clustering. Partitions Scanned shows how many partitions the engine actually read for the query, while Partitions Total shows how many partitions exist in the table. The relation between these two numbers reveals pruning efficiency: a small number of partitions scanned relative to the total means strong pruning, reading only the necessary data; a scanned count close to the total indicates more data is being read.

This isn’t about how many warehouses are used or the size of each partition. It’s about how many partitions were read versus how many exist, which is why the best interpretation is partitions read by the query compared to the total partitions in the table.