Date: Tue, 17 Mar 2020 14:56:16 -0500
From: Justin Pryzby <pryzby@telsasoft.com>
To: Laurenz Albe <laurenz.albe@cybertec.at>
Cc: Andres Freund <andres@anarazel.de>,
	Masahiko Sawada <masahiko.sawada@2ndquadrant.com>,
	David Rowley <dgrowleyml@gmail.com>,
	Tomas Vondra <tomas.vondra@2ndquadrant.com>,
	Amit Kapila <amit.kapila16@gmail.com>,
	Masahiko Sawada <sawada.mshk@gmail.com>,
	Darafei =?utf-8?Q?Kom=D1=8Fpa?= Praliaskouski <me@komzpa.net>,
	pgsql-hackers@lists.postgresql.org,
	Alvaro Herrera <alvherre@2ndquadrant.com>,
	Michael Banck <mbanck@gmx.net>
Subject: Re: Berserk Autovacuum (let's save next Mandrill)
Message-ID: <20200317195616.GZ26184@telsasoft.com>
References: 
 <CAApHDvrACF7jB_hbq75XiaRtFRUq9BcXm=f15icu3+Q9xD+YQA@mail.gmail.com>
 <CA+fd4k5HwScjcUbjSuZXWyU6JNENsOebktdi1ft39rJqLZ9N2A@mail.gmail.com>
 <20200316124713.GH26184@telsasoft.com>
 <9c9b7d19d17ad5e4e0ee10b8426fa1b78dd24c0a.camel@cybertec.at>
 <20200316201315.dgkeedl6rp2wdjje@alap3.anarazel.de>
 <19be41199e7b303bbecc1066bd9f011d95882aa8.camel@cybertec.at>
 <20200316213459.kh6xln3z35hgjodd@alap3.anarazel.de>
 <7c132215f0e947af7d9ccb7a340ad6e81b369c24.camel@cybertec.at>
 <20200317152409.GW26184@telsasoft.com>
 <e0b452925b00e2712e299a3370f85ee6be32889c.camel@cybertec.at>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <e0b452925b00e2712e299a3370f85ee6be32889c.camel@cybertec.at>
User-Agent: Mutt/1.9.4 (2018-02-28)
Precedence: bulk

On Tue, Mar 17, 2020 at 08:42:07PM +0100, Laurenz Albe wrote:
> Also, since aggressive^H^H^H^H^H^H^H^H^H^Hproactive freezing seems to be a
> performance problem in some cases (pages with UPDATEs and DELETEs in otherwise
> INSERT-mostly tables), I have done away with the whole freezing thing,
> which made the whole patch much smaller and simpler.
> 
> Now all that is introduced are the threshold and scale factor and
> the new statistics counter to track the number of inserts since the last
> VACUUM.
> 
> Updated patch attached.
> 
> Perhaps we can reach a consensus on this reduced functionality.

+1

I still suggest scale_factor maximum of 1e10, like
4d54543efa5eb074ead4d0fadb2af4161c943044

Which alows more effectively disabling it than a factor of 100, which would
progress like: ~1, 1e2, 1e4, 1e6, 1e8, 1e10, ..

I don't think that 1e4 would be a problem, but 1e6 and 1e8 could be.  With
1e10, it's first vacuumed when there's 10billion inserts, if we didn't previous
hit the n_dead threshold.

I think that's ok?  If one wanted to disable it up to 1e11 tuples, I think
they'd disable autovacuum, or preferably just implement an vacuum job.

The commit message says:
|The scale factor defaults to 0, which means that it is
|effectively disabled, but it offers some flexibility
..but "it" is ambiguous, so should say something like: "the table size does not
contribute to the autovacuum threshold".

-- 
Justin