Re: Removing the pgstat_flush_io() call from the walwriter - Mailing list pgsql-hackers

From Nazir Bilal Yavuz
Subject Re: Removing the pgstat_flush_io() call from the walwriter
Date
Msg-id CAN55FZ1VP6xUZSGLamZFXHtVcEA2UX9BSsA8UMhE5UGE9Qm58A@mail.gmail.com
Whole thread Raw
In response to Re: Removing the pgstat_flush_io() call from the walwriter  (Bertrand Drouvot <bertranddrouvot.pg@gmail.com>)
Responses Re: Removing the pgstat_flush_io() call from the walwriter
List pgsql-hackers
Hi,

On Thu, 19 Dec 2024 at 08:50, Bertrand Drouvot
<bertranddrouvot.pg@gmail.com> wrote:
>
> Hi,
>
> On Wed, Dec 18, 2024 at 11:55:11AM -0500, Andres Freund wrote:
> > Yea, i think it's fine to just call it unnecessarily. Particularly because we
> > do want to actually report IO stats for walwriter eventually.
>
> Yeah, better to spend time and energy on making it "necessary" instead: I'll
> try to have a look at adding IO stats for walwriter.

I have been working on showing all WAL stats in the pg_stat_io [1]
view but currently it is blocked because the size of the WAL read IO
may vary and it is not possible to show this on the pg_stat_io view.
Proposed a fix [2] for that (by counting IOs as bytes instead of
blocks in the pg_stat_io) which you already reviewed :).

[1] https://commitfest.postgresql.org/51/4950/
[2] https://commitfest.postgresql.org/51/5256/

-- 
Regards,
Nazir Bilal Yavuz
Microsoft



pgsql-hackers by date:

Previous
From: Greg Sabino Mullane
Date:
Subject: Re: Send duration output to separate log files
Next
From: Bertrand Drouvot
Date:
Subject: Re: Removing the pgstat_flush_io() call from the walwriter