Warning: This blog is about to delve back into the arcana of IP address governance. This may seem strange, as most of the Internet policy world is buzzing with high-level discussion of Secretary Clinton’s latest Internet Freedom speech and the role of Facebook and Twitter in the Arab world revolutions. Those are great topics, but we remain steadfastly fixated upon the boring, nitty-gritty technical details of how the Internet is governed, both because no one else covers them and because those details probably will, in the long run, have more impact on Internet freedom than politicians’ speeches.
Our last post noted the looming battle over IP address Whois. Since that came out a policy proposal has been placed before ARIN that could be construed as an attempt to lay the groundwork for a more open and competitive market in post-allocation IP address registration services. The proposal is called “ARIN-prop-133: No Volunteer Services on Behalf of Unaffiliated Address Blocks.” It was put before the ARIN Public Policy Mailing List on 13 February by Benson Schliesser, a Cisco employee (the usual disclaimers about him acting as an individual apply).
The great thing about this proposal is that it finally faces a crucial fact about the evolution of IP address governance institutions and, instead of agonizing and hand-wringing about it, actually tries to do something. The central fact of which I speak is that a very large portion of the ipv4 address space (about 40%) consists of what the community calls “legacy holders” – i.e., people who received their IPv4 address blocks before ARIN existed and thus are not under any contract with ARIN. Under our current system, ARIN maintains records of those allocations, and those records purport to be authoritative. But in many cases the organizations that actually hold and control those address blocks are not members of ARIN.
By the way, Syracuse University is one of those legacy holders. I believe we hold one or more /16s. And I can’t get our network administrators the least bit interested in becoming an ARIN member.
The author of ARIN-prop-133 finds this situation troublesome, and makes a good case for why it is:
“…absent an agreement such as the RSA [Registry Services Agreement] or LRSA [Legacy RSA] there is no formal relationship with legacy address holders. At present, however, ARIN continues to provide services to these organizations. This is done without compensation and potentially in opposition to the legacy address holders' wishes. As a result of this behavior ARIN has created an illusion of implied authority that exposes ARIN to unacceptable levels of liability, is hindering the development of an open address market (driving it “underground”), and is putting the operational stability of the Internet at risk. As new services such as RPKI are contemplated this situation becomes even more critical.”
The phrase “illusion of implied authority” hits the nail on the head. ARIN-prop-133 responds to this dilemma by proposing to force us to shit or get off the pot. It says that we should no longer pretend that ARIN “serves” legacy address block holders and is the authoritative registry for all of them when there is neither a contractual relationship between the parties nor any indication from the address block holder that it even wants ARIN to provide those services for it. It says that ARIN should cease providing “all directory services, reverse mapping services, and [any] future services that may be provided to the community” unless it is specifically asked to do so by the relevant address block holder.
According to the proposal, organizations that want ARIN to provide services must sign one of its contracts, known as a Registration Services Agreement (RSA). For organizations that do not opt-in to receive registry services, ARIN is limited to providing no more than a pointer (such as a RWhois referral) to an authoritative directory service for that holder's legacy address blocks. In other words, this proposal is explicitly founded on the assumption that many of the services currently provided by ARIN are independent of its function as the initial allocator of address blocks, and could be provided competitively – provided that the Whois directory is integrated and consistent across all providers.
Given the hidebound conservatives who dominate ARIN’s PPML and Board, there is little chance that ARIN-prop-133 will be passed any time soon. But it is a carefully thought out proposal drafted by someone who knows what he is doing, and thus cannot simply be ignored or dismissed. There are those in ARIN who believe that this is a larger issue that must be addressed within ICANN.