Re: Todo: Teach planner to evaluate multiple windows in the optimal order

On Wed, Feb 15, 2023 at 3:02 PM David Rowley <dgrowleyml@gmail.com> wrote:
>
> On Wed, 15 Feb 2023 at 17:23, John Naylor <john.naylor@enterprisedb.com> wrote:
> > HEAD:
> >
> > 4 ^ 8: latency average = 113.976 ms
> > 5 ^ 8: latency average = 783.830 ms
> > 6 ^ 8: latency average = 3990.351 ms
> > 7 ^ 8: latency average = 15793.629 ms
> >
> > Skip rechecking first key:
> >
> > 4 ^ 8: latency average = 107.028 ms
> > 5 ^ 8: latency average = 732.327 ms
> > 6 ^ 8: latency average = 3709.882 ms
> > 7 ^ 8: latency average = 14570.651 ms

> Yeah, just a hack. My intention with it was just to prove we had a
> problem because sometimes Sort -> Incremental Sort was faster than
> Sort.  Ideally, with your change, we'd see that it's always faster to
> do the full sort in one go. It would be good to see your patch with
> and without the planner hack patch to ensure sort is now always faster
> than sort -> incremental sort.

Okay, here's a rerun including the sort hack, and it looks like incremental sort is only ahead with the smallest set, otherwise same or maybe slightly slower: