[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [aso-policy] RIR proposal
Dear Bill,
dear Community,
please let me start with a statement of sincere recognition,
support and appreciation for what IANA has accomplished in the
past and what's going to be added to it's track record ion the
future!
My comments as attached are in no way meant to insult anyone
(as a non-native speaker there's always a potential for misusing
words and a danger of embarrassing readers without intention and
even without knowing it.)
As an introduction, I feel it is appropriate to briefly review
history (as good as I can - being a new kid on the block), (my)
assumptions, and recent developments (which I had the privilege
to observe or be involved with). Read on with a grain of salt.
IANA was set up and has got a chance to grow in line with, and
in support of, the development of the Internet as we know it
today. While it was continuously growing and changing, it was
still some sort of a very stable and recognized entity, along-
side the development of IETF (and friends), the establishment
of ISPs (and commercial service providers, eventually) and the
birth of the hierarchical and distributed registry model.
This stability was based on and did involve a few aspects:
- continuous availability of staff that was recognized as an
authority to apply to in case of decisions to be taken
- funding stability
- "natural feedback loops" with a well-known peer community,
including the IETF, the Regional Registries, and the other
"implicitely recognized" players in the field.
I think many would still take all of those things for granted,
and "we" still tend to align our postions to fit that model and
those assumptions.
But, from my point of view, these assumption do no longer apply,
at least not all of them. Why?
- As one of the biggest blows, we lost Jon,
and we're seeing other more recent changes in IANA staff.
- There were fundamental changes in IANA's funding structure,
and, most importantly,
- we saw the establishment of the RIRs, quite a few years ago,
and
- we saw the establishment of ICANN and the Supporting
Organisations.
Ok, after giving you my background on the topic at hand, I'll
now go to your comments (which are very much appreciated!).
=% There are of course examples in the past where IANA allocations have
=% been made in a framework where consequent through would suggest
=% an alternative action. Bill note advocates IANA as a place where
=% 'reasonable requests' can be made, predicated that in order for
=% a request to reach IANA the entire RIR structure would've had to
=% have been subverted in some way in order to get to IANA in any case.
=
= "subversion" is not the issue here. what is important
= is the fact that fundamental changes in the way addressing
= is managed have occured with some frequency in the past
= and there is no reason to believe this will not continue
= into the future.
This is indeed a prudent assumption.
= In the evolution of IPv4, two major shifts
= included the original imposition of the the subnet model
= (classes) and then the imposition of CIDR. Both these changes
= were derived from outside the existent RIR nee IR process
= and were tested in the engineering community with direct
= access to the IANA for addressing changes.
While this (i.e. deriving) might be true [to some degree, only!]
I do not see why the same goals, and in particualr the testing,
could (should??) not be accomplished while working with the
RIRs' constituencies, i.e. the ISPs. What am I missing?
=% What is NOT at issue here is the ability of ICANN to set global
=% policy for the RIRs - what IS at issue is that it is then
=% most unseemly for IANA to have some independet ability of
=% allocation OUTSIDE OF THESE VERY POLICIES THAT ICANN WANTS
=% TO IMPOSE ON THE RIRs my emphasis). Of course if the response
=% were to be that the allocations would conform to these policies,
=% then there would be no problem in making the allocation through
=% the RIR structure in the first place, right?
=
= This is a capability that the IANA has always had. The
= proposal is asking the IANA to abdicate this flexablity
= to the RIRs.
Yes, the "old" IANA always has had that "right" and flexibility.
As to my knowledge, this right was only exercised in very rare
circumstances, and after consulting IANA's "natural" peers.
However, I'm of the opinion that this "old" IANA is no longer
existing. Whatever IANA is going to evolve into, has to be
considered as a part/segment/activity/branch of ICANN.
Putting that thought into words makes me recognize even more
painfully than before that I'm not aware of a published
desription or charter for ICANN within the IANA frame-work.
Any pointers to such a document would be very helpful for me.
In addition to the more formal and policy-related issues,
there's an additional, and more down-to-earth aspect as well:
keeping track of address "delegations".
Now, let me use the RIR terminology which is more precise,
I think. Keeping track of IP address allocations and IP
assignments, as well as of AS-Numbers for that matter, is one of
the basic responsibilites and services provided by the Regional
Registries (aka The Adress Registry). This, in turn is (more or
less seamlessly) linked to the components of the Routing
Registry.
So, to summarize, I do not see IANA being any longer the good
old traditional, completely "independent" and "ultimately
responsible" entity in a position to conduct it's business (with
regard to IPv[46] addresses and AS#s and protocol identifiers)
"outside" or "alongside" the established ICANN+SO and RIR
frame-work. Quite to the contrary, I do perceive IANA as an
integral part of ICANN.
And ICANN is a fairly new set-up, with lots of new people in key
positions which *do* have their individual wisdom to contribute,
their strengths and determination to work for and in the best
interest of the Internet. But at the same time, this is no
longer the "old IANA". And we have already heard about proposals
and/or transactions within ICANN/IANA involving address space,
which I think should not be possible without external scrutiny.
However, there was a valid point made with regard to technical
tests and pilot projects, like Net38 and the IPv6 pTLA
environment.
I agree that the proposed policy document could benefit from
clearly defining what address custodianship by ICANN/IANA stands
for, what ICANN's/IANA's role with regard to supporting any new
technical proposals or solutions would be and what the on-going
collaboration with established technical bodies (like the IETF
*and* the RIR constituencies) should look like.
=% I would like also to note that much thought has gone into this
=% proposal, and that thought is based on the RIR policies and
=% processes to date, and based on the acceptance of the ICANN
=% structure into the future. These policies reflect what we have
=% learned about address resource administration in the Internet, and
=% how we manage to strike a balance between the myriad of interests
=% that are at play in this environment. To claim, as Bill does, that
=% IANA somehow needs the reserve ability to exercise resource allocation
=% based on its ability to make hasty decisions with no visible rationale
=% (synonyms for 'quick action' and 'long term vision' I believe)
=% is one which I do not believe is in the best interests of the
=% Internet, nor in the best interests of its panopoly of interested
=% parties.
=
= True, much thought has gone into the proposal and I don;t
= think it should be abandoned. I do believe that the policies
= reflect the RIR lessons of the commodity Internet and it is
= reasonable for ICANN to take these types of proposals under
= serious consideration. After all, ICANN is a very new organisation.
= However it is prudent to note that the whole RIR structure
= is of recent vintage as well. My particular RIR is about
= three years old and the others are roughly the same. The IANA
= task has had about two decades of experience.
Yes, the *old* IANA has had about two decades of experience!
But at the same time I'm worried about what the IANA is going
to look like in the (near) future, and what it's operational
environment would be, eventually.
= So yes, I do think it is wise to allow the IANA to retain
= its ability to delegate as we have seen in the definition
= and early deployment of the 6bone, before the RIRs had any
= understanding of the various issues.
I can follow and support the lines of thought up to the
technical aspects, e.g. early deployment, test-beds, and the
like, but I cannot agree to the phrase "before the RIRs had any
understanding of the various issues".
This wording would suggest that the RIRs have an independent life
of their own, without any involvement of and feedback from the
technical, ISP and user communities. While I cannot comment on
the ARIN and/or APNIC region, I'd say this is not true for the
ripe region. *Not at all*!!!
= There are times when
= a decision must be made -AND- documented when there is no
= "visible rationale". The net 39 experiment was one recent
= example. This would not have happened within the construct
= of the RIR proposal and without it having been managed by the
= IANA, the Internet would be much worse off now.
I don't agree that it would have been impossible to accomplish
the same result while involving the RIRs and/or their
constituencies!
I'd even venture to say that doing it "the other way 'round"
would help to avoid the need for taking decisions without a
"visible rationale".
But again, I read that as mixing technology development and
support for new ideas with address allocations and assignments
*for production use*.
= I'll agree
= that the IANA retaining its discresionary capability is
= a potent tool and that it is reasonable to ask that its
= use receive careful scrutiny. It is unreasonable to abdicate
= that capability because the "panopoly" of commercial interests
= wish a stable future. In a bad light, this could be viewed as
= an attempt by the RIRs to restrict entry into the address
= registration arena.
Sorry, I can't follow the logic in that paragraph!?
What do you propose? To solve a "restricting entry into the
address registration arena" problem by adding *one* registry,
i.e. IANA to compete with the RIRs?
Or do you propose to treat IP addresses and AS numbers like
domain names and create a competitive market for the sale of
address space? I'm lost....
=% Geoff Huston
=%
=% Member, Executive Committee, APNIC
=
= Are you speaking for the APNIC EC?
=
=--bill
=* on-line archive: http://aso.icann.org/wilma-bin/wilma/aso-policy *
=* To unsubscribe: send "unsubscribe" to aso-policy-request@aso.icann.org *
=--------------------------------------------------------------------------------
My comments so far
[ as an AC member (not speaking for any RIR, AC or ASO),
the RIPE Database Working Group chair
and for ACOnet's Local-IR. ]
Wilfried
--------------------------------------------------------------------------
Wilfried Woeber : e-mail: Woeber@CC.UniVie.ac.at
Computer Center - ACOnet : Tel: +43 1 4277 - 140 33
Vienna University : Fax: +43 1 4277 - 9 140
Universitaetsstrasse 7 : RIPE-DB Handle: WW144
A-1010 Vienna, Austria, Europe : PGP public key ID 0xF0ACB369
__________________________________________________________________________
* on-line archive: http://aso.icann.org/wilma-bin/wilma/aso-policy *
* To unsubscribe: send "unsubscribe" to aso-policy-request@aso.icann.org *
|