From: Antonin Houska <ah@cybertec.at>
To: Andres Freund <andres@anarazel.de>
Cc: pgsql-hackers@postgresql.org, Noah Misch <noah@leadboat.com>,
    Heikki Linnakangas <hlinnaka@iki.fi>,
    Robert Haas <robertmhaas@gmail.com>,
    Thomas Munro <thomas.munro@gmail.com>
Subject: Re: AIO writes vs hint bits vs checksums
In-reply-to: <2305.1727247975@antos>
References: <stj36ea6yyhoxtqkhpieia2z4krnam7qyetc57rfezgk4zgapf@gcnactj4z56m>
 <2305.1727247975@antos>
Comments: In-reply-to Antonin Houska <ah@cybertec.at>
   message dated "Wed, 25 Sep 2024 09:06:15 +0200."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2738.1727251947.1@antos>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 25 Sep 2024 10:12:27 +0200
Message-ID: <2739.1727251947@antos>
Archived-At: <https://www.postgresql.org/message-id/2739.1727251947%40antos>
Precedence: bulk

Antonin Houska <ah@cybertec.at> wrote:

> Andres Freund <andres@anarazel.de> wrote:
> =

> > What I'd instead like to propose is to implement the right to set hint=
 bits as
> > a bit in each buffer's state, similar to BM_IO_IN_PROGRESS. Tentativel=
y I
> > named this BM_SETTING_HINTS. It's only allowed to set BM_SETTING_HINTS=
 when
> > BM_IO_IN_PROGRESS isn't already set and StartBufferIO has to wait for
> > BM_SETTING_HINTS to be unset to start IO.
> > =

> > Naively implementing this, by acquiring and releasing the permission t=
o set
> > hint bits in SetHintBits() unfortunately leads to a significant perfor=
mance
> > regression. While the performance is unaffected for OLTPish workloads =
like
> > pgbench (both read and write), sequential scans of unhinted tables reg=
ress
> > significantly, due to the per-tuple lock acquisition this would imply.
> =

> An alternative approach: introduce a flag that tells that the checksum i=
s
> being computed, and disallow setting hint bits when that flag is set. As=
 long
> as the checksum computation takes take much less time than the IO, fewer=
 hint
> bit updates should be rejected.

Well, the checksum actually should not be computed during the IO, so the I=
O
would still disallow hint bit updates :-(

-- =

Antonin Houska
Web: https://www.cybertec-postgresql.com