Re: moving system catalogs to another tablespace

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Alvaro Herrera <alvherre(at)commandprompt(dot)com>
Cc: Jaime Casanova <jcasanov(at)systemguards(dot)com(dot)ec>, Euler Taveira de Oliveira <euler(at)timbira(dot)com>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: moving system catalogs to another tablespace
Date: 2009-10-06 14:58:11
Message-ID: 1699.1254841091@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> I don't agree that pg_largeobject is a random catalog either -- it is,
> in fact, the only catalog in which an interesting size is to be
> expected.

Yeah, I have sometimes thought that pg_largeobject shouldn't be
considered a system catalog at all. It's more nearly like a toast
table, ie, it's storing "out of line" user data.

This has some interesting connections with the proposed changes
for associating privilege data with large objects. The proposed
"meta" table would certainly qualify as a system catalog still.
Would there be any sense in redefining pg_largeobject as an actual
toast table attached to that catalog? Probably not, or at least
it wouldn't directly contribute to solving Jaime's problem.
But it seems like now would be a good time to think outside the
box a little bit about where we want to go with pg_largeobject.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Simon Riggs 2009-10-06 14:58:47 Re: Hot Standby on git
Previous Message Robert Haas 2009-10-06 14:49:46 Re: Patch: create or replace language