Re: Does a GRANT on a table cascade/implied to its SEQUENCES - Mailing list pgsql-general

From Roderick A. Anderson
Subject Re: Does a GRANT on a table cascade/implied to its SEQUENCES
Date
Msg-id 44C66167.10609@acm.org
Whole thread Raw
In response to Re: Does a GRANT on a table cascade/implied to its SEQUENCES  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
Tom Lane wrote:
> "Roderick A. Anderson" <raanders@acm.org> writes:
>
>>I've add a new user and as I go though granting various accesses to the
>>different tables I realized many of those tables have primary keys that
>>are generated by a sequence.
>
>
>>Do I need to grant access on the sequences and what type of access --
>>SELECT for sure but what about UPDATE -- for each of the tables I've
>>granted the user access to?
>
>
> Right now, GRANT on a table doesn't do anything about subsidiary
> sequences.  (There have been discussions about changing that, but
> nothing's happened yet.)  So if you want someone to be able to INSERT
> into a table with a serial column, you need to give them UPDATE rights
> on the sequence.  Offhand I see no direct reason why they'd need SELECT
> rights on the sequence, but maybe they do.

Thanks for the clarification Tom.


Rod
--

pgsql-general by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: Plperl return_next and bytea binary data?
Next
From: "Philippe Lang"
Date:
Subject: Re: Plperl return_next and bytea binary data?