Date: Wed, 31 Mar 2021 09:16:09 -0400
From: Bruce Momjian <bruce@momjian.us>
To: Vineet Naik <naikvin@gmail.com>
Cc: pgsql-admin@lists.postgresql.org
Subject: Re: Finding out if 'vacuum --analyze-in-stages' has generated enough
 stats
Message-ID: <20210331131609.GB11609@momjian.us>
References: 
 <CADmbCiMDXsimr2b-S7HKWyOiVOWTazDhABEassUqA+e4z8UMmg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CADmbCiMDXsimr2b-S7HKWyOiVOWTazDhABEassUqA+e4z8UMmg@mail.gmail.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Archived-At: 
 <https://www.postgresql.org/message-id/20210331131609.GB11609%40momjian.us>
Precedence: bulk

On Wed, Mar 31, 2021 at 12:38:14PM +0530, Vineet Naik wrote:
> > This option will try to create some statistics as fast as possible, to make the database usable, and then produce full statistics in the subsequent stages.
> 
> So my question is, after which stage is it ok to send traffic to the
> server without much performance impact? Also, is it possible to query
> the stats view in pg and determine if enough stats have been
> generated?

That is a very hard question to answer.  Some queries have only a few
possible plans, so few or no statistics are fine.  Other queries require
detailed statistics to generate efficient plans.  I think you need to
set up a dummy server with no stats and try your workload on that, then
run the stages manually and test after each stage to see how the
performance is.

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

  If only the physical world exists, free will is an illusion.