Re: bug in DROP TABLESPACE - Mailing list pgsql-hackers

From Tom Lane
Subject Re: bug in DROP TABLESPACE
Date
Msg-id 1623.1089424725@sss.pgh.pa.us
Whole thread Raw
In response to Re: bug in DROP TABLESPACE  (Christopher Kings-Lynne <chriskl@familyhealth.com.au>)
Responses Re: bug in DROP TABLESPACE  (Bruce Momjian <pgman@candle.pha.pa.us>)
Re: bug in DROP TABLESPACE  (Christopher Kings-Lynne <chriskl@familyhealth.com.au>)
List pgsql-hackers
Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes:
> The other thing we need are these two commands:

> ALTER DATABASE foo SET TABLESPACE spc;
> ALTER SCHEMA foo SET TABLESPACE spc;

> I think these should not be considered new features but essential 
> functionality left out of the original patch.

The latter is relatively trivial as long as you interpret it as just
changing the default TS for tables created later in that schema.
(On the other hand, you can fake that with a simple UPDATE to
pg_namespace, so I'm not sure it qualifies as a "must have".)

The former is a real mess since it would require moving some of the
system catalogs.  (At least the nailed-in-cache ones, but I'm not sure
we've yet determined which ones *have* to be in the database's default
tablespace.)  It'll be fairly entertaining to move pg_class in
particular, but I doubt you can build such a thing out of the available
spare parts at all.

My vote would be to postpone any such patch to 7.6 (or 8.1 as the case
may be), because it's a nontrivial addition of functionality and we had
no prototype as of July 1.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: More vacuum.c refactoring
Next
From: Alvaro Herrera
Date:
Subject: Re: Nested Transactions, Abort All