Re: catalog access with reset GUCs during parallel worker startup

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Andres Freund <andres(at)anarazel(dot)de>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: catalog access with reset GUCs during parallel worker startup
Date: 2022-02-10 02:47:09
Message-ID: CA+TgmoavOVOOncezNvWW+CH6HeWAMj+GHr5F_13=KBKzvSsAdQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, Feb 9, 2022 at 7:33 PM Andres Freund <andres(at)anarazel(dot)de> wrote:
> > Can we arrange to absorb the leader's values before starting the
> > worker's transaction, instead of inside it?
>
> I assume Robert did it this way for a reason? It'd not be surprising if there
> were a bunch of extensions assuming its safe to do catalog accesses if
> IsUnderPostmaster or such :(

I remember trying that, and if I remember correctly, it broke with
core GUCs, without any need for extensions in the picture. I don't
remember the details too well, unfortunately, but I think it had to do
with the behavior of some of the check and/or assign hooks.

It's probably time to try it again, because (a) maybe things are
different now, or (b) maybe I did it wrong, and in any event (c) I
can't really remember why it didn't work and we probably want to know
that.

--
Robert Haas
EDB: http://www.enterprisedb.com

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Julien Rouhaud 2022-02-10 02:54:30 Re: Database-level collation version tracking
Previous Message Michael Paquier 2022-02-10 02:43:32 Re: Logging in LockBufferForCleanup()