Re: Changing the setting of wal_sender_timeout per standby

From: Michael Paquier <michael(at)paquier(dot)xyz>
To: "Tsunakawa, Takayuki" <tsunakawa(dot)takay(at)jp(dot)fujitsu(dot)com>
Cc: 'Masahiko Sawada' <sawada(dot)mshk(at)gmail(dot)com>, Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>, "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Changing the setting of wal_sender_timeout per standby
Date: 2018-09-19 06:13:03
Message-ID: 20180919061303.GB19808@paquier.xyz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, Sep 19, 2018 at 02:40:31PM +0900, Michael Paquier wrote:
> Parameters classified as PGC_BACKEND can be updated by any users, and
> those marked as PGC_SU_BACKEND can only be updated by superusers.
> Replication users are not superusers, which is why PGC_BACKEND is most
> adapted. Your description should just say "This parameter cannot be
> changed after session start.

Actually, now that I look at guc.c, using PGC_BACKEND breaks one case:
such parameter types cannot be changed even with SIGHUP, and this even
if the session relies on the default value. So you could break
applications relying on reloads. PGC_USERSET would actually do the work
as if the session uses a non-default value specified by SET or within
the connection string, then SIGHUP updates have no effect. On the
contrary, if the client relies on the default value, then SIGHUP updates
take effect. Sorry for the confusion, I should have looked at guc.c
more carefully.
--
Michael

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tsunakawa, Takayuki 2018-09-19 06:21:52 RE: Changing the setting of wal_sender_timeout per standby
Previous Message Edmund Horner 2018-09-19 06:04:43 Re: Tid scan improvements