Mr. Jon Postel
University of Southern California
As I am sure you are aware, TRUE Root Name Servers are intended to
provide stability to the Internet. RFC 2010 has many good ideas about
how to harden these servers. All of the other TRUE Root Name Servers
that I am aware of are following RFC 2010 to the letter. Are the new
servers you describe, 2010 compliant ?
In your note below you point out that these four machines are temporarily
housed at ISI, which I assume means the University of Southern California.
In the U.S. Government's InterNIC file [shown below], it indicates that
two of the machines are at NSI which I assume is Network Solutions, Inc.
Can you be more specific ?
A traceroute to one of the machines goes via LN.NET, an ISP that you run.
The leg before that is via "genuity" where you are on the Board of Directors
of this Bechtel funded company <http://www.genuity.net>. Are LN.NET
or Bechtel involved in this project ?
As more TLD Registries come on line, they will depend on the various
confederations of TRUE Root Name Servers for service. There will be
many reasons why ISPs and companies select the confederation they
use. As long as each confederation refers users to the proper TLD
Name Servers, downstream caching is coherent because the TLD
name servers do most of the real work.
Having said this, I am curious whether you view your NEW Root Name
Servers to be purely for researchers to use or whether you intend
broader use? I note that you mention this is an "experiment". Are
you concerned that ISPs might use these servers and experience
operational integrity problems ? Should ISPs be cautious about that ?
Have you been following the discussions on the NANOG list regarding
some of the problems they have had with the legacy Root Name Servers ?
Since everyone's goal is to make sure the Internet remains
stable and grows in an organized way that does not deny service
to people around the world, I think that actions taken in the
arena of TRUE Root Name Servers need to be done carefully.
As I am sure you are aware, once ISPs adopt certain servers
they rarely change and they follow the U.S. Government's lead.
I also note that you are using the U.S. Government's InterNIC
file distribution system to easily facilitate the wide-spread adoption
of these "experimental" servers. Can you explain who at the
National Science Foundation (NSF) authorized that action ?
In light of the fact that some mail lists seem to filter information
and some now appear to be deleting postings from their archives.
I am posting this to several groups that I feel will be interested
and involved in these discussions.
Thanks for your time.