Re: regarding statistics retaining with 18 Upgrade - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: regarding statistics retaining with 18 Upgrade
Date
Msg-id aRyYBRIJW0PBH7De@momjian.us
Whole thread Raw
In response to Re: regarding statistics retaining with 18 Upgrade  (Fujii Masao <masao.fujii@gmail.com>)
List pgsql-hackers
On Tue, Nov 18, 2025 at 12:41:28PM +0900, Fujii Masao wrote:
> On Tue, Nov 18, 2025 at 8:56 AM Bruce Momjian <bruce@momjian.us> wrote:
> >
> > On Sat, Nov 15, 2025 at 06:52:15PM -0500, Robert Treat wrote:
> > > On Sat, Nov 15, 2025 at 9:32 AM Bruce Momjian <bruce@momjian.us> wrote:
> > > > On Sat, Nov 15, 2025 at 06:02:53PM +0800, Rambabu V wrote:
> > > > > Hi Team,
> > > > >
> > > > > seems stats are not retained after upgrade. Below are the artifacts , is it
> > > > > expected? as per the release notes statistics should be retained.
> > > >
> > > > Uh, it is the _optimizer_ statistics that are preserved by pg_upgrade,
> > > > not the cumulative statistics system:
> > > >
> > > >         https://www.postgresql.org/docs/current/monitoring-stats.html
> > > >
> > > > Yes, they are easily confused.
> > > >
> > >
> > > I wonder if it might be worth trying to clarify this a little bit
> > > more... perhaps something like the attached?
> >
> > Patch applied back to PG 18, thanks.
> 
> Thanks for working on this!
> 
>    <para>
> -   If <option>--statistics</option> is specified,
> +   When <option>--statistics</option> is specified,
>     <command>pg_dumpall</command> will include most optimizer statistics in the
> -   resulting dump file.  However, some statistics may not be included, such as
> -   those created explicitly with <xref linkend="sql-createstatistics"/> or
> -   custom statistics added by an extension.  Therefore, it may be useful to
> +   resulting dump file. This does not include all statistics, such as
> +   those created explicitly with <xref linkend="sql-createstatistics"/>,
> +   custom statistics added by an extension, or statistics collected by the
> +   cumulative statistics system.  Therefore, it may still be useful to
>     run <command>ANALYZE</command> on each database after restoring from a dump
>     file to ensure optimal performance.  You can also run <command>vacuumdb -a
>     -z</command> to analyze all databases.
> 
> Since pgupgrade.sgml contains similar text, I think it should be
> updated as well.
> For example, something like this:

Yeah, you are probably right.  I didn't see that.  You should apply it
since it is your patch, thanks.

-- 
  Bruce Momjian  <bruce@momjian.us>        https://momjian.us
  EDB                                      https://enterprisedb.com

  Do not let urgent matters crowd out time for investment in the future.



pgsql-hackers by date:

Previous
From: Fujii Masao
Date:
Subject: Allow GUC settings in CREATE SUBSCRIPTION CONNECTION to take effect
Next
From: Robert Haas
Date:
Subject: Re: pg_plan_advice