Re: [RFC] Extend namespace of valid guc names

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Hannu Krosing <hannu(at)2ndQuadrant(dot)com>
Cc: Andres Freund <andres(at)2ndQuadrant(dot)com>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: [RFC] Extend namespace of valid guc names
Date: 2013-09-06 21:36:59
Message-ID: 6625.1378503419@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hannu Krosing <hannu(at)2ndQuadrant(dot)com> writes:
> On 09/06/2013 08:48 PM, Tom Lane wrote:
>> Well, sure, but I would think that ALTER SYSTEM SET should be constrained
>> to only set known GUCs, not invent new ones on the fly.

> What's the reasoning behind this ?

If you don't know what a GUC is, you don't know what are valid values for
it, and thus you might write an illegal value into auto.conf (or whatever
we're calling it this week). That could have consequences as bad as
failure to restart, should the DBA decide to preload the module defining
that GUC, which would then complain about the bad value during postmaster
start.

> I was assuming that ALTER SYSTEM SET would allow all GUCs which
> do not require restart which includes all "newly invented" ones.

I do not believe that the former need imply the latter, nor do I see a
strong use-case for allowing ALTER SYSTEM SET on session-local GUCs,
which is what any truly invented-on-the-fly GUCs would be. The whole
business with session-local GUCs is pretty much a kluge anyway, which
we might want to retire or redefine someday; so I'd much prefer that
ALTER SYSTEM SET stayed out of it.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2013-09-06 21:47:11 Re: Fix picksplit with nan values
Previous Message Greg Stark 2013-09-06 20:55:09 Re: Valgrind Memcheck support