On Tue, Nov 16, 2021 at 3:42 PM Andres Freund <andres@anarazel.de> wrote:
> On 2021-11-16 15:19:19 -0500, Robert Haas wrote:
> > > Hm. I think this might included a bunch of convoluting factors that make it
> > > hard to judge the actual size of the performance difference.
> >
> > Yes, I think so, too.
>
> FWIW I ran that pgench thing I presented upthread, and I didn't see any
> meaningful and repeatable performance difference 354a1f8d220, ad26ee28250 and
> 0002 applied ontop of ad26ee28250. The run-to-run variance is way higher than
> the difference between the changes.
Thanks. I suspected that the results I was seeing were not meaningful,
but it's hard to be sure when the results seem to be repeatable
locally.
I'm still not entirely clear on whether you prefer v1-0002, v2-0002,
or something else.
--
Robert Haas
EDB: http://www.enterprisedb.com