From: Tom Lane <tgl@sss.pgh.pa.us>
To: David Rowley <dgrowleyml@gmail.com>
cc: Nathan Bossart <nathandbossart@gmail.com>, 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.
In-reply-to: 
 <CAApHDvrcM4G0==52jFoAvmxk6ZRjXKq7nkik=fqq6BH=dSJsYA@mail.gmail.com>
References: <19086-871ff11017d03dc5@postgresql.org> <aO6mUj9H2WpEFylG@nathan>
 <1000041.1760471415@sss.pgh.pa.us> <aO7IoIEnQDbtFzrs@nathan>
 <CAApHDvoU2O-_zKNL-toHwkDcMXCzzvgRUTnA3baEEhCB4vqMRw@mail.gmail.com>
 <aO_KwnHm2AVn0E4m@nathan>
 <CAApHDvrcM4G0==52jFoAvmxk6ZRjXKq7nkik=fqq6BH=dSJsYA@mail.gmail.com>
Comments: In-reply-to David Rowley <dgrowleyml@gmail.com>
	message dated "Thu, 16 Oct 2025 10:38:09 +1300"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1324002.1760565328.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 15 Oct 2025 17:55:28 -0400
Message-ID: <1324003.1760565328@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1324003.1760565328%40sss.pgh.pa.us>
Precedence: bulk

David Rowley <dgrowleyml@gmail.com> writes:
> Seems to be due to pg_get_indexdef / pg_get_constraintdef operating on
> a cold cat cache. Getting rid of those the rewritten version runs in
> 1.8 seconds with 100k tables for me.

I wonder how much win could be had by postponing those function calls
so that they only act on indexes we're going to dump.  It might be
a net loss in the default dump-everything case, though.

Also, it looks to me like getIndexes does not even look at the result
of pg_get_constraintdef unless contype =3D=3D 'x'.  So there should be
some low-hanging fruit with

-                         "pg_catalog.pg_get_constraintdef(c.oid, false) A=
S condef, "
+                         "CASE WHEN c.contype =3D 'x' THEN "
+                         "pg_catalog.pg_get_constraintdef(c.oid, false) "
+                         "END AS condef, "

This wouldn't matter except with primary/unique constraints, but
surely there are plenty of those in a typical DB.

			regards, tom lane