Re: "Write once-Read many" table ?

From: Niall Litchfield <>
Date: Sun, 2 Mar 2008 06:07:39 -0800 (PST)
Message-ID: <>

On Feb 28, 4:23 pm, wrote:
> On Feb 28, 2:46 pm, Mladen Gogala <> wrote:
> > On Thu, 28 Feb 2008 03:17:19 -0800, jm.scheiwiler wrote:
> > >> You have to trust your DBA. That's the bottom line.
> > >> --
> > > The thing is, I am the dba.
> > > I trust myself but that's not enough
> > So, who else, besides you, can do grants & revokes?
> > --
> > Mladen Gogala
> In fact, business management want to be sure that a line inserted will
> never be changed, by anyone.
> They would like a feature to ensure that.
> Something that would be guaranteed by construction in the database,
> guaranteed by oracle ...
> I ask the question on their behalf.

In the end they are unlikely to like the answer which is that you cannot be sure (in the sense of certain) that a line will never be changed. There are features- such as database vault or the use of write-once media which can increase confidence but they do not deliver guarantees.

Database Vault can be relatively trivially sidestepped by someone with access to the oracle software owner's account on the server (create a new password file) and may not be available for your version, writeonce  media tends not to be written to at the time of the initial write (a new piece of media for every single write is unsustainable) instead people move data to it periodically, triggers can be circumvented and so on.

Security and audit are both important and need to be addressed on a number of levels not least the human level. Incidentally I'd love to see the law in question that imposes this perceived requirement, it would be a nonsensical and non-implementable law, but I rather suspect that it doesn't exist.

Niall Litchfield Received on Sun Mar 02 2008 - 08:07:39 CST

Original text of this message