[anti-abuse-wg] [db-wg] RIPE NCC's proposed implementation of Abuse Contact Management in the RIPE Database

Kaveh,

On Thursday, 2012-11-15 13:30:09 +0100, 
Kaveh Ranjbar <kranjbar at ripe.net> wrote:
> Abuse handling is generally a role within an organisation and the
> RIPE NCC is planning to model this relation in the RIPE Database. So
> the "abuse-c:" attribute will be available in the ORGANISATION object
> and should reference a ROLE object. When a ROLE object is referenced
> from an ORGANISATION object by the "abuse-c:" attribute, it must have
> an "abuse-mailbox:" attribute.
> 
> All resources allocated or assigned by the RIPE NCC are required to
> have a reference to an ORGANISATION object in the RIPE Database. By
> adding a single "abuse-c:" attribute to their ORGANISATION object, a
> resource holder can quickly cover all their resources. All resources
> that are allocated or assigned by the RIPE NCC and which are subject
> to and compliant with current resource policies will then either
> directly reference an abuse contact or will inherit one from the
> parent resource.
> 
> The detailed changes to RIPE Database objects and business rules,
> fine tuning, search options and facilities for quick data entry by
> RIPE NCC members using the LIR Portal are all described in an article
> on RIPE Labs:
> 
> https://labs.ripe.net/Members/kranjbar/implementation-details-of-policy-2011-06

I can't get to that article as the site is currently down, but basically
I think you're describing something like this:


+----------+  org:  +--------------+
| resource +------->| ORGANISATION |
+----------+        +--------+-----+
                             |
                             | abuse-c:
                             v
 --------                  +------+
( e-mail )<----------------+ ROLE |
 --------   abuse-mailbox: +------+

While it seems a bit complicated, the layers of indirection make sense,
and if the Whois server follows the chains automatically then it should
be workable.

A maintainer only has to create a single mailbox and point to it once.

A user gets the correct e-mail spit out somewhere logical in their
Whois query.

Cool. Looking forward to having this implemented so we can move on to
the much scarier phase of abuse contact management. :)

Cheers,

--
Shane