From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
cc: Nathan Bossart <nathandbossart@gmail.com>,
        Melanie Plageman <melanieplageman@gmail.com>,
        Thomas Munro <thomas.munro@gmail.com>,
        Michael Paquier <michael@paquier.xyz>,
        Fujii Masao <masao.fujii@oss.nttdata.com>,
        "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Subject: Re: race condition when writing pg_control
In-reply-to: <20240516185805.hnbt2tr23q7xbx3m@awork3.anarazel.de>
References: <20200527071053.GD103662@paquier.xyz>
 <fc796148-7d63-47bb-e91d-e09b62a502e9@oss.nttdata.com>
 <A9F111A3-560A-4C21-A35B-9D3182704187@amazon.com>
 <20200602052427.GC3594@paquier.xyz>
 <CA+hUKGJJJOP1jfw4yRqozXghiqtXRH0gisHyNwa+nzCL_xyjcg@mail.gmail.com>
 <20200603020259.GA89559@paquier.xyz>
 <CA+hUKGKToqjdeRtr8ZHVbPToShun5ifjioj6x3nsjseOkhCMEQ@mail.gmail.com>
 <CAAKRu_YNGwEYrorQYza_W8tU+=toXRHG8HpyHC-KDbZqA_ZVSA@mail.gmail.com>
 <20240516183558.GA1620903@nathanxps13> <2136775.1715885450@sss.pgh.pa.us>
 <20240516185805.hnbt2tr23q7xbx3m@awork3.anarazel.de>
Comments: In-reply-to Andres Freund <andres@anarazel.de>
	message dated "Thu, 16 May 2024 11:58:05 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2138497.1715886091.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Thu, 16 May 2024 15:01:31 -0400
Message-ID: <2138498.1715886091@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2138498.1715886091%40sss.pgh.pa.us>
Precedence: bulk

Andres Freund <andres@anarazel.de> writes:
> On 2024-05-16 14:50:50 -0400, Tom Lane wrote:
>> The intention was certainly always that it be atomic.  If it isn't
>> we have got *big* trouble.

> We unfortunately do *know* that on several systems e.g. basebackup can r=
ead a
> partially written control file, while the control file is being
> updated.

Yeah, but can't we just retry that if we get a bad checksum?

What had better be atomic is the write to disk.  Systems that can't
manage POSIX semantics for concurrent reads and writes are annoying,
but not fatal ...

			regards, tom lane