パーティションプルーニング

The Snowy Peak data team runs the following query against the subscriptions table, which holds three years of billing data across hundreds of micro-partitions. Snowflake stores min and max start_date values for every micro-partition in the table.

What does Snowflake do with this metadata when the query runs?

SELECT COUNT(*)
FROM subscriptions
WHERE start_date = '2024-03-15';