Re: Berserk Autovacuum (let's save next Mandrill)

From: Laurenz Albe <laurenz(dot)albe(at)cybertec(dot)at>
To: David Rowley <dgrowleyml(at)gmail(dot)com>
Cc: Andres Freund <andres(at)anarazel(dot)de>, Alexander Korotkov <a(dot)korotkov(at)postgrespro(dot)ru>, Justin Pryzby <pryzby(at)telsasoft(dot)com>, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>, Masahiko Sawada <masahiko(dot)sawada(at)2ndquadrant(dot)com>, Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>, Darafei Komяpa Praliaskouski <me(at)komzpa(dot)net>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Michael Banck <mbanck(at)gmx(dot)net>
Subject: Re: Berserk Autovacuum (let's save next Mandrill)
Date: 2020-03-28 04:12:11
Message-ID: 61d1ab62a22ee7976b8d099dda8c8089fae74a9f.camel@cybertec.at
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Sat, 2020-03-28 at 11:59 +1300, David Rowley wrote:
> On Fri, 27 Mar 2020 at 22:40, Laurenz Albe <laurenz(dot)albe(at)cybertec(dot)at> wrote:
> > The new meaning of -2 should be documented, other than that it looks
> > good to me.
>
> But the users don't need to know anything about -2. It's not possible
> to explicitly set the value to -2. This is just the reset value of the
> reloption which means "use the GUC".

I see.

> > I'll accept the new semantics, but they don't make me happy. People are
> > used to -1 meaning "use the GUC value instead".
>
> The problem with having -1 on the reloption meaning use the GUC, in
> this case, is that it means the reset value of the reloption must be
> -1 and we need to allow them to set -2 explicitly, and if we do that,
> then -1 also becomes a valid value that users can set. Maybe that's
> not the end of the world, but I'd rather have the reset value be
> unsettable by users. To me, that's less confusing as there are fewer
> special values to remember the meaning of.
>
> The reason I want a method to explicitly disable the feature is the
> fact that it's easy to document and it should reduce the number of
> people who are confused about the best method to disable the feature.
> I know there's going to be a non-zero number of people who'll want to
> do that.

In the light of that, I have no objections.

Yours,
Laurenz Albe

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2020-03-28 04:22:51 Re: error context for vacuum to include block number
Previous Message Guancheng Luo 2020-03-28 03:28:41 Re: [PATCH] Check operator when creating unique index on partition table