Comparing Query Performance when using an “ideal” Nonclustered Index versus Columnstore Index
I gave a presentation on columnstore indexing last week and one question I received was regarding the performance of a (hypothetical) narrow, supporting nonclustered index