Re: Grouped Index Tuples

Jim C. Nasby wrote:
> On usage, ISTM it would be better to turn on GIT only for a clustered
> index and not the PK? I'm guessing your automatic case is intended for
> SERIAL PKs, but maybe it would be better to just make that explicit.

Not necessarily; since often (in my tables at least) the data for
come columns has some local grouping of similar values even though
it's not the clustered index.

The obvious example is addresses (city, state, street, etc may not appear
clustered - but if you cluster your table by zip code, GIT will work
well for them).

Other examples would be tables containing dates and product names - even
though there's no total monotonic ordering if your product families
get refreshed every couple years you'll find that one range of product
names shows up mostly in old years, and others in new years - so GIT
could prove useful there - despite not being a clustered index.