Feedback-ID: i0fe9450f:Fastmail
Date: Tue, 16 Sep 2025 15:44:25 +0900
From: Michael Paquier <michael@paquier.xyz>
To: Kirill Reshke <reshkekirill@gmail.com>
Cc: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>,
	Robert Haas <robertmhaas@gmail.com>,
	pgsql-hackers@lists.postgresql.org
Subject: Re: relfilenode statistics
Message-ID: <aMkHSeitgUPtMyA2@paquier.xyz>
References: <Zt/ZeIeWq0UVSAuy@ip-10-97-1-34.eu-west-3.compute.internal>
 <ZyiTiqWPbRMseQjb@ip-10-97-1-34.eu-west-3.compute.internal>
 <CA+TgmoZtRxvAuMyO4bQjj5t47XWvQsowPAedkOkrCP61B7mQFg@mail.gmail.com>
 <Zym1xlMhuzPdekMt@ip-10-97-1-34.eu-west-3.compute.internal>
 <CALdSSPhQyvK_wsvV+HiQcC6f57Zg2WO60wxv=CXbcoZ1p5HHEQ@mail.gmail.com>
 <Z0nbomp2e/L30WDA@ip-10-97-1-34.eu-west-3.compute.internal>
 <CALdSSPhFaLEWXG9Mn8-tZoS68rE52647a2oZ6jrUNSBtwadD0A@mail.gmail.com>
 <Z07d850Yp7Sd9Qqq@ip-10-97-1-34.eu-west-3.compute.internal>
 <Z3gN3bwUMijyV4I0@ip-10-97-1-34.eu-west-3.compute.internal>
 <CALdSSPh76=ACN=z68oM44ZMgZsQxTTYVF8qvi1F6TGA5Mn6uiQ@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
	protocol="application/pgp-signature"; boundary="vnrT+5yAGHMoTbKl"
Content-Disposition: inline
In-Reply-To: 
 <CALdSSPh76=ACN=z68oM44ZMgZsQxTTYVF8qvi1F6TGA5Mn6uiQ@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/aMkHSeitgUPtMyA2%40paquier.xyz>
Precedence: bulk


--vnrT+5yAGHMoTbKl
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Mar 13, 2025 at 02:00:52PM +0500, Kirill Reshke wrote:
> Hmm. While it is true that catalog lookups cannot be performed during
> crash recovery, is it really necessary to save and retrieve statistics
> after a crash?

Yes, losing stats on crash is a *very* annoying thing.  Having no
stats for a relation means that autovacuum gives up entirely on
relations it has no stats of, skipping it entirely until they have
rebuilt and bloat would accumulate.  Being able to recover these stats=20
=66rom crash recovery is a cheap design, that would improve reliability
by a large degree.

> Given that statistics are permitted to be outdated and
> server crashes are anticipated to be infrequent, it looks loke losing
> a few analysis runs due to server crashes is acceptable.
> In any case, I am totally OK with the relfilenode-based method because
> it is generally less restricted (to other postgresql parts e.g. wal-
> replay ) and simpler.

The startup process is not connected to a database and has no access
to pg_class: the only thing we can know about are the on-disk files,
not their in-catalog OIDs.  FWIW, I think that this patch would be a
huge step forward a more reliable stats system.

True that the patch needs a rebase.  Bertrand has also mentioned that
some points needed more work.
--
Michael

--vnrT+5yAGHMoTbKl
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----
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=Zptb
-----END PGP SIGNATURE-----

--vnrT+5yAGHMoTbKl--