Re: Planning time of Generic plan for a table partitioned into a lot

From: David Rowley <david(dot)rowley(at)2ndquadrant(dot)com>
To: Amit Langote <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp>
Cc: "Kato, Sho" <kato-sho(at)jp(dot)fujitsu(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Planning time of Generic plan for a table partitioned into a lot
Date: 2018-11-30 06:57:09
Message-ID: CAKJS1f_Y1cZX68tKdO28NiMbO8dzr=Q8dskRa+oR6E1dRHfnDg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Fri, 30 Nov 2018 at 19:44, Amit Langote
<Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp> wrote:
> Yeah, maybe we haven't explained in the documentation where generic plans
> are described that making them for partitioned table is an expensive
> affair. Although, by definition, they are built once for a given query
> and PG 11 with it's execution-time pruning can execute these plans pretty
> quickly, which is an overall improvement. But you'd obviously know that
> much. :)

Maybe a documents patch is a good idea. Your planner patches for
partitioning are only going to widen the performance gap between
generating a plan where many partitions are pruned vs generating a
generic plan where no partition could be pruned. I can only imagine
this will cause problems for more people as that gap widens as it will
likely lead to more and more people with partitioned tables with high
numbers of partitions.

--
David Rowley http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Kyotaro HORIGUCHI 2018-11-30 07:05:54 Re: idle-in-transaction timeout error does not give a hint
Previous Message Noah Misch 2018-11-30 06:51:40 Re: [HACKERS] Weaker shmem interlock w/o postmaster.pid