Snowflake tries to auto-sort its containers to fit the query pattern; most of these platforms use at least one data ordering. Vertica allows multiple projection

Snowflake tries to auto-sort its containers to fit the query pattern; most of these platforms use at least one data ordering. Vertica allows multiple projections on a table with different orderings and distribution keys. On top of that BRIN (block range) indexes are usually used to capture the value of sorting by pruning I/O. I don't see these mentioned here -- they seem like a good open-source version of this idea.