Date: Wed, 18 Jan 2023 11:49:19 -0600
From: Justin Pryzby <pryzby@telsasoft.com>
To: Laurenz Albe <laurenz.albe@cybertec.at>
Cc: Bruce Momjian <bruce@momjian.us>,
	Nathan Bossart <nathandbossart@gmail.com>,
	Tomas Vondra <tomas.vondra@enterprisedb.com>,
	Robert Haas <robertmhaas@gmail.com>,
	=?iso-8859-1?Q?=C1lvaro?= Herrera <alvherre@alvh.no-ip.org>,
	yuzuko <yuzukohosoya@gmail.com>, pgsql-hackers@postgresql.org
Subject: Re: document the need to analyze partitioned tables
Message-ID: <20230118174919.GA9837@telsasoft.com>
References: <19c248d9-3687-9e0f-0b1a-2adad1de6fbe@enterprisedb.com>
 <CA+TgmoYNaSWQg1W3JZ62UBM9iMphBpVJr_oDxn5T1X_viOPBtw@mail.gmail.com>
 <20220315230011.GZ28503@telsasoft.com>
 <88903179-5ce2-3d4d-af43-7830372bdcb6@enterprisedb.com>
 <1fd81ddc7710a154834030133c6fea41e55c8efb.camel@cybertec.at>
 <20230112232747.GA2111950@nathanxps13>
 <Y8cKxCRXBwEVFfot@momjian.us>
 <20230117210050.GX9837@telsasoft.com>
 <Y8cQJIMFAe7QT73/@momjian.us>
 <fae7e128fd78965a1fa4dc9c571a9f5187d16f85.camel@cybertec.at>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <fae7e128fd78965a1fa4dc9c571a9f5187d16f85.camel@cybertec.at>
User-Agent: Mutt/1.9.4 (2018-02-28)
Archived-At: 
 <https://www.postgresql.org/message-id/20230118174919.GA9837%40telsasoft.com>
Precedence: bulk

On Wed, Jan 18, 2023 at 10:15:18AM +0100, Laurenz Albe wrote:
> On Tue, 2023-01-17 at 16:16 -0500, Bruce Momjian wrote:
> > On Tue, Jan 17, 2023 at 03:00:50PM -0600, Justin Pryzby wrote:
> > > Maybe (all?) the clarification the docs need is to say:
> > > "Partitioned tables are not *themselves* processed by autovacuum."
> > 
> > Yes, I think the lack of autovacuum needs to be specifically mentioned
> > since most people assume autovacuum handles _all_ statistics updating.

That's what 61fa6ca79 aimed to do.  Laurenz is suggesting further
clarification.

> > Can someone summarize how bad it is we have no statistics on partitioned
> > tables?  It sounds bad to me.
> 
> Andrey Lepikhov had an example earlier in this thread[1].  It doesn't take
> an exotic query. 
> 
> Attached is a new version of my patch that tries to improve the wording.

I tweaked this a bit to end up with:

> -    Partitioned tables are not processed by autovacuum.  Statistics
> -    should be collected by running a manual <command>ANALYZE</command> when it is
> +    The leaf partitions of a partitioned table are normal tables and are processed
> +    by autovacuum; however, autovacuum does not process the partitioned table itself.
> +    This is no problem as far as <command>VACUUM</command> is concerned, since
> +    there's no need to vacuum the empty, partitioned table.  But, as mentioned in
> +    <xref linkend="vacuum-for-statistics"/>, it also means that autovacuum won't
> +    run <command>ANALYZE</command> on the partitioned table.
> +    Although statistics are automatically gathered on its leaf partitions, some queries also need
> +    statistics on the partitioned table to run optimally.  You should collect statistics by
> +    running a manual <command>ANALYZE</command> when the partitioned table is
>      first populated, and again whenever the distribution of data in its
>      partitions changes significantly.
>     </para>

"partitions are normal tables" was techically wrong, as partitions can
also be partitioned.

-- 
Justin