Re: POC: GROUP BY optimization

Alexander Korotkov <aekorotkov@gmail.com> writes:
> On Wed, Dec 27, 2023 at 5:23 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> I think it's a fool's errand to even try to separate different sort
>> column orderings by cost.

> Besides sorting column orderings by cost, this patch also tries to
> match GROUP BY pathkeys to input pathkeys and ORDER BY pathkeys.  Do
> you think there is a chance for the second part if we leave the cost
> part aside?

I think it's definitely reasonable to try to match up available
orderings, because that doesn't really require fine distinctions
of cost: either it matches or it doesn't.  Eliminating a sort step
entirely is clearly a win.  (Incremental sort complicates this though.
I doubt our cost model for incremental sorts is any good either, so
I am not eager to rely on that more heavily.)

            regards, tom lane