Re: Schema (namespace) privilege details - Mailing list pgsql-hackers

From Sander Steffann
Subject Re: Schema (namespace) privilege details
Date
Msg-id 000901c1e861$25d9d870$64c8a8c0@balefire10ww
Whole thread Raw
In response to Re: Schema (namespace) privilege details  (Curt Sampson <cjs@cynic.net>)
Responses Re: Schema (namespace) privilege details  (Curt Sampson <cjs@cynic.net>)
List pgsql-hackers
Hi,

> Curt Sampson <cjs@cynic.net> writes:
> > On Fri, 19 Apr 2002, Sander Steffann wrote:
> >> I can't think of a reason that [creation of] temp tables should
> >> be prevented.
>
> > Maybe to keep hostile users from filling up your disk?
>
> That does come to mind --- but if you've let hostile users into
> your database, filling your disk is not exactly the smallest problem
> they could cause.  They can very easily cause DOS problems just based
> on overconsumption of CPU cycles, or on crashing your server constantly.
> (Cm'on, we all know that can be done.)  Even more to the point, is there
> nothing in your database that you'd not want published to the entire
> world?  There's got to be a certain amount of trust level between you
> and the persons you allow SQL-command-level access to your database.
> If not, you ought to be interposing another level of software.
>
> My current proposal for schema protection does include a TEMP-table-
> creation right ... but to be honest I am not convinced that it'd be
> worth the trouble to implement it.  Comments anyone?

I see your point, but I think Curt is right... If users are always allowed
to make temp tables, you can't give someone real read-only access to the DB.
I agree that there has to be more protection to prevent other abuses, but at
least the disk is safe.

Sander




pgsql-hackers by date:

Previous
From: "Rod Taylor"
Date:
Subject: Re: Foreign keys and indexes.
Next
From: "."@babolo.ru
Date:
Subject: Re: [INTERFACES] sqlbang