On Sun, Mar 15, 2020 at 10:38 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Well, if we say "PG can handle up to 64K relations in a query",
> I think people would read that as meaning that you actually get
> usable performance with up to 64K relations. Which is a long
> way away, even if certain specific cases might work acceptably.
> The existing docs discourage using more than a few thousand
> partitions, IIRC, and that seems like sufficient guidance for now.
Would would be the downside to raising INNER_VAR to, say, a billion? I
mean, I get that queries involving many tables aren't necessarily
going to work very well, but due to all of the work that has been done
in this area over the last couple of years, they have a much better
chance of working well now than they did before, and I think there is
some appetite to continue fixing things that would let them work even
better in the future.
Incidentally, this is the second complaint I've seen about this
problem in the last week, so it seems pretty clear that the
partitioning work is actually moving the needle in this area. We ought
to be happy, not sad, that people are trying to use more partitions;
we ought to be receptive, not resistant, to their reports of trouble;
and if we decide that a particular problem is absolutely impossible to
fix, we should certainly not resist documenting those restrictions
which are unavoidable. Refusing to give people either a fix or any
useful information about the limits seems very unhelpful.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company