ICANN ICANN Email List Archives

[draft-ssr-role-remit]


<<< Chronological Index >>>    <<< Thread Index >>>

Coordination with root server operators

  • To: <draft-ssr-role-remit@xxxxxxxxx>
  • Subject: Coordination with root server operators
  • From: Roberto Gaetano <roberto_gaetano@xxxxxxxxxxx>
  • Date: Thu, 30 Aug 2012 23:30:33 +0300

IMHO, the most interesting issue raised is the coordination with root server
operators, I am therefore concentrating my attention to this item.

For the security, stability and reliability (SSR) of the Internet, the
coordination of the root server operators, and their collaboration, is of
the paramount importance.

The keyword here is "accountability". If a framework is set up to improve
Internet SSR, who has to be held responsible (and therefore who has to be
given the authority to make decisions)?

So far, the root server operators have done an excellent job. So far,
questions like "transitioning a root server" have never been an issue. Also,
historically the root server operators have kept their independence against
any attempt from third parties (including governments and ICANN) to exert
any "coordination" role.

However, with the fast development of the Internet (in terms of number of
users but also in terms of commercial value), we need to think globally, and
be "future proof". So, the main question hidden in this report is whether
the current root server operators coordination mechanism (excellent up to
now and in the foreseeable future) is the optimal solution in the long term.

I leave to technical gurus the answer to questions whether the current
technical developments are sufficient to guarantee Internet SSR, and
concentrate on the "governance" issue underlying the question.

Are the current root server operators "accountable" to the global Internet
community (for SSR issues)? My straight answer is "no".

Is ICANN "accountable" to the global Internet community (for SSR issues)? My
straight answer is "not yet".

The major difference is that the root server operators, great technical guys
to whom I pay immense respect, rely on a "Jon Postel model", that is based
on technical excellence and straight moral attitude. In the same way I
believed back in the late 90's that a transition was needed for the DNS, I
now believe that a transition is needed for the root server system. Not
because the current operators are not technically able to do the task, or do
not have the high moral standard to be independent from commercial
pressures, but simply because there is no formal accountability mechanism,
and the system relies mostly, if not only, on the good will of individuals.

Now back to the point. What is ICANN's role?

My vision is that ICANN fully embraces the multistakeholder approach. The
consensus of the stakeholders is what gives ICANN the accountability, and
therefore the authority to coordinate these issues. Some might argue, and
they might be not completely wrong, that ICANN is not there yet. However,
"we" (as ICANN is "all of us") are getting there.

An ICANN that is *fully* a multistakeholder-based organization (fully
inclusive geopolitically, balancing commercial and noncommercial interests,
balancing supplier and consumer interests, establishing - and *enforcing* -
contracts that are fair to the whole community) will have the authority to
assume the role of coordinator of the root server operations, managing the
transition should an operator cease, but also to enforce rules that improve
Internet SSR, like DNSSEC.

Let's make a simple example. I am just an Internet user. If technical
improvements to the Internet SSR are needed to improve my user experience, I
would demand to have a forum in which I have a voice to express my concerns,
comments, needs, wishes, and what else. And a governance model that ensures
that my voice will be not only heard, but also taken into account.

I do not have this at all with the current root server operators governance
model. I am not 100% happy with ICANN model either, but I do believe that
the latter can be improved. And I have even a mechanism to improve the model
itself - although I reckon that it might take time.

Cheers,

R.

 

 



<<< Chronological Index >>>    <<< Thread Index >>>

Privacy Policy | Terms of Service | Cookies Policy