Re: PG15 beta1 sort performance regression due to Generation context change

On 7/15/22 6:52 PM, Andres Freund wrote:
> Hi,
>
> On 2022-07-15 18:40:11 -0400, Jonathan S. Katz wrote:
>> What I find interesting is the resistance to adding any documentation around
>> this feature to guide users in case they hit the regression. I understand it
>> can be difficult to provide guidance on issues related to adjusting
>> work_mem, but even just a hint in the release notes to say "if you see a
>> performance regression you may need to adjust work_mem" would be helpful.
>> This would help people who are planning upgrades to at least know what to
>> watch out for.
>
> If we want to add that as boilerplate for every major release - ok, although
> it's not really actionable. What I'm against is adding it specifically for
> this release - we have stuff like this *all the time*, we just often don't
> bother to carefully find the specific point at which an optimization might
> hurt.
>
> Honestly, if this thread teaches anything, it's to hide relatively minor
> corner case regressions. Which isn't good.

I think it's OK to discuss ways we can better help our users. It's also
OK to be wrong; I have certainly been so plenty of times.

Jonathan