On Wed, Mar 27, 2019 at 08:15:35PM -0700, Andres Freund wrote:
>Hi,
>
>On 2019-03-27 21:13:46 +0100, Tomas Vondra wrote:
>> On Wed, Mar 27, 2019 at 12:32:34PM -0700, Peter Geoghegan wrote:
>> > On Wed, Mar 27, 2019 at 12:04 PM Andres Freund <andres@anarazel.de> wrote:
>> > > Congrats, this was long in the making!
>> >
>> > +1
>> >
>>
>> Thanks! I guess it's a Gordie Howe of our patches.
>
>Btw, do you have plans to work on using more of this information during
>join planning? Or at least an estimate of how much work that'd be?
>
Yes, I'd like to use more of this info in join estimation. I don't have
an exact plan how to do that - so if anyone has ideas regarding this,
I'd love to hear them.
An obvious option is to add statistics covering multiple tables (the
syntax is already designed with that in mind). But I have a feeling there
are things we can do without that, e.g. using multivariate MCV lists in
eqjoinsel roughly the same way we use per-column MCV lists, etc.
regards
--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services