On Monday, February 28, 2011 9:15:27 am Tom Lane wrote:
> Adrian Klaver <adrian(dot)klaver(at)gmail(dot)com> writes:
> > I understand, yet I don't:) Seems I have a misconception of the the
> > FOREIGN KEY process. The error was on the query below, which I am
> > taking is the query you refer to above. To me it looks like a look up
> > from the referencing(permout) table to the referenced(out2cp) one. The
> > part I am not clear on is why that runs with the permissions of the
> > referenced table not the referencing table?
> Because we force it that way so that you don't have to splatter
> permissions all around to support a foreign key check. Generally
> the internal FK check queries will only fail if the owner of a table
> revokes his own permissions, which isn't too usual.
> regards, tom lane
Had not thought about it from that angle. Thanks.
In response to
pgsql-general by date
|Next:||From: Alan Acosta||Date: 2011-02-28 17:43:58|
|Subject: Lock ACCESS EXCLUSIVE and Select question !|
|Previous:||From: Tom Lane||Date: 2011-02-28 17:15:27|
|Subject: Re: Permission denied when inserting |