Re: ALTER TABLE ... SET TABLESPACE pg_default

From: Craig Ringer <craig(at)2ndquadrant(dot)com>
To: Stephen Frost <sfrost(at)snowman(dot)net>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: ALTER TABLE ... SET TABLESPACE pg_default
Date: 2014-01-20 02:33:14
Message-ID: 52DC8AEA.7090507@2ndquadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 01/17/2014 05:28 AM, Stephen Frost wrote:
> Greetings,
>
> Harking back to 10 years ago when tablespaces were added, it looks
> like we originally figured that users didn't need permissions to
> create tables in the database default, per 2467394e. That strikes
> me as perfectly fair. Unfortunately, the later addition of ALTER
> TABLE ... SET TABLESPACE (af4de814) didn't get the memo about the
> default tablespace being special in this regard and refuses to let
> a user move their tables into the default tablespace, even though
> they can do so via 'CREATE TABLE ... AS SELECT * FROM ...'.
>
> Barring objections, I'll add the same conditional around the
> AclCheck in ATPrepSetTableSpace() as exists in DefineRelation() to
> allow users to ALTER TABLE ... SET TABLESPACE into the database's
> default tablespace and backpatch accordingly.

Sounds sensible. I just stumbled across a report of this bug, too:

http://stackoverflow.com/questions/21193127/avoid-users-to-create-tables-on-default-tablespace

--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Florian Pflug 2014-01-20 02:40:37 Re: [PATCH] Negative Transition Aggregate Functions (WIP)
Previous Message Craig Ringer 2014-01-20 02:05:48 Re: currawong is not a happy animal