Message-ID: <78846bd91600d804e6befa72c4a6ff4e7c522410.camel@j-davis.com>
Subject: Re: Proposal to add page headers to SLRU pages
From: Jeff Davis <pgsql@j-davis.com>
To: "Li, Yong" <yoli@ebay.com>
Cc: Aleksander Alekseev <aleksander@timescale.com>, PostgreSQL Hackers
 <pgsql-hackers@lists.postgresql.org>, "Bagga, Rishu" <bagrishu@amazon.com>,
  Robert Haas <robertmhaas@gmail.com>, "Debnath, Shawn" <sdn@ebay.com>,
 Andrey Borodin <x4mmm@yandex-team.ru>,  "Shyrabokau, Anton"
 <antons@ebay.com>
Date: Thu, 14 Mar 2024 15:27:24 -0700
In-Reply-To: <51E39019-D805-4EB9-A814-EBC63C43223C@ebay.com>
References: 
	<CAJ7c6TMUY=H2Hn=YbdYimk_NfVGq1QR9UhaPsE2Zd+8rgzdZMg@mail.gmail.com>
	 <IA1PR07MB982925731C4C0B38DC8F2F7FB98BA@IA1PR07MB9829.namprd07.prod.outlook.com>
	 <F10F07A0-19E9-49AD-8832-6936F8984274@yandex-team.ru>
	 <162841701969726@mail-sendbernar-corp-main-71.iva.yp-c.yandex.net>
	 <76110763-F29E-469E-AA62-2EED17875BB7@ebay.com>
	 <CA+TgmoZ2i5smJuGDRqsUbEKHga_xOc27f6=59ci9aLjA0c2_KA@mail.gmail.com>
	 <DM4PR07MB9812DA12DA729E8C709DEDD9B98AA@DM4PR07MB9812.namprd07.prod.outlook.com>
	 <1404E248-F2B2-48BC-9B5E-BB318F3BE583@amazon.com>
	 <628CFBEF-DF56-4D9A-A2F1-0A417848B6B6@ebay.com>
	 <CAJ7c6TMrFrUF1W3nO=w+3mJu_G8UaQkMJh2aV9WhP20Xs0M_6g@mail.gmail.com>
	 <359BB3DC-A174-43AF-A9BA-064CB6D2C7E9@ebay.com>
	 <48CDC195-5030-4426-9E02-931D8AE14DF7@ebay.com>
	 <4B689BA2-270F-4B26-AAA2-65BB17C03AC4@ebay.com>
	 <3ee88744547f6865fe2f48b2b6f058a50796cb91.camel@j-davis.com>
	 <51E39019-D805-4EB9-A814-EBC63C43223C@ebay.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu2 
MIME-Version: 1.0
Archived-At: <https://www.postgresql.org/message-id/78846bd91600d804e6befa72c4a6ff4e7c522410.camel%40j-davis.com>
Precedence: bulk

On Mon, 2024-03-11 at 10:01 +0000, Li, Yong wrote:
> - The clog LSN group has been brought back.
> =C2=A0 Now the page LSN on each clog page is used for honoring the write-
> ahead rule
> =C2=A0 and it is always the highest LSN of all the LSN groups on the page=
.
> =C2=A0 The LSN groups are used by TransactionIdGetStatus() as before.

I like where this is going.

=C3=81lvaro, do you still see a problem with this approach?

> - New comments have been added to pg_upgrade to mention the SLRU
> =C2=A0 page header change as the reason for upgrading clog files.

That seems reasonable, but were any alternatives discussed? Do we have
consensus that this is the right thing to do?

And if we use this approach, is there extra validation or testing that
can be done?

Regards,
	Jeff Davis