Date: Wed, 15 Oct 2025 11:24:34 -0500
From: Nathan Bossart <nathandbossart@gmail.com>
To: David Rowley <dgrowleyml@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, andrewbille@gmail.com,
	pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #19086: pg_dump --data-only selects and do not uses index
 definitions for the dumped tables.
Message-ID: <aO_KwnHm2AVn0E4m@nathan>
References: <19086-871ff11017d03dc5@postgresql.org>
 <aO6mUj9H2WpEFylG@nathan>
 <1000041.1760471415@sss.pgh.pa.us>
 <aO7IoIEnQDbtFzrs@nathan>
 <CAApHDvoU2O-_zKNL-toHwkDcMXCzzvgRUTnA3baEEhCB4vqMRw@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAApHDvoU2O-_zKNL-toHwkDcMXCzzvgRUTnA3baEEhCB4vqMRw@mail.gmail.com>
Archived-At: <https://www.postgresql.org/message-id/aO_KwnHm2AVn0E4m%40nathan>
Precedence: bulk

On Wed, Oct 15, 2025 at 12:56:25PM +1300, David Rowley wrote:
> I tried rewriting the query so that pg_attribute is joined to rather
> than subqueries. With 1500 tables I get:
> 
> master:
> 
> ignore_system_indexes = on
> Execution Time: 6853.262 ms
> 
> ignore_system_indexes = off
> Execution Time: 66.781 ms
> 
> Rewritten query:
> 
> ignore_system_indexes = on
> Execution Time: 53.351 ms
> 
> ignore_system_indexes = off
> Execution Time: 56.965 ms

I tried this with 100K tables and saw the following (ignore_system_indexes
= off):

master:
 Planning Time: 1.672 ms
 Execution Time: 4077.008 ms

rewritten:
 Planning Time: 1.641 ms
 Execution Time: 3427.206 ms

I tried this with ignore_system_indexes = on, too, but it was very slow.

-- 
nathan