Re: Get rid of pgstat_count_backend_io_op*() functions

From: Bertrand Drouvot <bertranddrouvot(dot)pg(at)gmail(dot)com>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org, Andres Freund <andres(at)anarazel(dot)de>
Subject: Re: Get rid of pgstat_count_backend_io_op*() functions
Date: 2025-09-02 15:25:39
Message-ID: aLcMc5FuyfeWVTuJ@ip-10-97-1-34.eu-west-3.compute.internal
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

On Tue, Sep 02, 2025 at 07:41:59AM +0000, Bertrand Drouvot wrote:
> On Tue, Sep 02, 2025 at 08:19:22AM +0900, Michael Paquier wrote:
> > Removing both the function calls and the extra counter incrementations
> > means to do the same thing as the WAL stats, where we have one
> > structure in charge of incrementing the counters, then the data diffs
> > are fed when flushing the entries in all the stats kinds. So this
> > would imply the introduction of an equivalent to WalUsage, but applied
> > to the IO stats, like a IOUsage, or something like that.
>
> Yeah, that sounds more complicated at first but I think that would solve
> the "ordering" issue mentioned above.

Attached is a draft version that implements an equivalent of WalUsage, that way:

- no ordering concerns anymore
- we can also get rid of PgStat_PendingIO ("replaced" by IOUsage)
- we can also get rid of PgStat_BackendPending (since both IO and WAL stats now
use the WAL's approach)

Regards,

--
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com

Attachment Content-Type Size
v2-0001-Get-rid-of-pgstat_count_backend_io_op-functions.patch text/x-diff 14.5 KB

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Nathan Bossart 2025-09-02 15:38:23 Re: Improve LWLock tranche name visibility across backends
Previous Message Jacob Champion 2025-09-02 15:18:42 Re: pgsql: oauth: Add unit tests for multiplexer handling