Re: Logging access to data in database table

From: Ivan Radovanovic <radovanovic(at)gmail(dot)com>
To: Misa Simic <misa(dot)simic(at)gmail(dot)com>
Cc: Bill Moran <wmoran(at)potentialtech(dot)com>, pgsql-general(at)postgresql(dot)org
Subject: Re: Logging access to data in database table
Date: 2012-01-25 19:37:56
Message-ID: 4F205A14.1040201@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 01/25/12 20:02, Misa Simic napisa:
> Thanks Bill,
>
> Make sense... db_link is probably then solution... Everything depends on
> concrete problem...
>
> But I still think security should be reconsidered (I would use db_link
> just in case there is no other options - if we must let users to have
> direct access to DB)... I mean, in that case when we need log each
> request for some sensitive data - we would not allow some user direct
> access to DB where he would be able to do such thing BEGIN TRAN, execute
> function what returns sensitive data, ROLLBACK Tran; (or many other
> things...)
>
> at least there would be an application layer above DB... (concretly in
> our case - Users do not have access to DB at all... everything is
> through Web App, actually DB - Web Service - User Apps (Web, Windows,
> Mobile etc...))
>
> Thanks,
>
> Misa
>
> 2012/1/25 Bill Moran <wmoran(at)potentialtech(dot)com
> <mailto:wmoran(at)potentialtech(dot)com>>
>
> In response to Misa Simic <misa(dot)simic(at)gmail(dot)com
> <mailto:misa(dot)simic(at)gmail(dot)com>>:
> >
> > But maybe it would be better to reorganise security on the way
> that users
> > who do not need to have access to some data - simply do not have it
> > (instead of to give them data and latter check log to confirm
> they have
> > taken it...)
>
> In many cases that's not enough. For example with HIPAA in the US,
> a user
> may be allowed to access data, but there still _has_ to be a log record
> for each access.
>
> --
> Bill Moran
> http://www.potentialtech.com
> http://people.collaborativefusion.com/~wmoran/
>
>
Thanks for reply,
Bill was right, this is security requirement that is independent of all
other security mechanisms we have implemented in this system :-)

I will check contrib/dblink - it seems to be one of the ways to solve
this problem

Best regards,
Ivan

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Adrian Klaver 2012-01-25 22:07:40 Re: Why extract( ... from timestamp ) is not immutable?
Previous Message Ivan Radovanovic 2012-01-25 19:32:17 Re: Logging access to data in database table