Why split IP allocations from the Internic?

Michael Dillon michael at MEMRA.COM
Wed Feb 26 15:54:09 EST 1997


On Wed, 26 Feb 1997, Jim Fleming wrote:

> What is wrong with the plan to clone the InterNIC many times
> that is well underway ?
> 
> Here is the short form...
> 	1. Clone the InterNIC 49 times

Compare the .US domain name registry with the .CA domain name registry and
you will immediately see the problem. Even though the two TLD's seem to be
subdivided similarily on the surface, they are not operated the same way
and that has made all the difference.

entity.city.state.US and entity.city.province.CA are commonly seen in 
both domains. But in the .US domain, there are 50 main registries
operating divided on state boundaries. This has led to coordination
problems, policy problems, public relations problems, etc... It's too
darn complicated and confusing. But the .CA domain is run by one single
national committee with one single national policy. There are
representatives from each province on the CA domain committee and it works
quite well, is efficient and has a good public image. 

As network engineers and network operators I think we all understand the
scaling problems that arise with a full mesh network. This is something
fundamental to the nature of network systems whether they are networks of
BGP peers or whether they are human organizations. ARIN is on the right
track by setting up a single central wide-area registry that will be
directly responsive to the members in its region and will only need to
coordinate activities and policies with 4 other regional registries (RIPE,
APNIC, AfricNIC and CORAS) as well as IANA.

Note: I just made up the names AfricNIC and CORAS (Consejo de Operadores
de la Red de America del Sur) and the real NIC's will probably figure
out even better names :-)

Michael Dillon                   -               Internet & ISP Consulting
Memra Software Inc.              -                  Fax: +1-250-546-3049
http://www.memra.com             -               E-mail: michael at memra.com




More information about the Naipr mailing list