CIADA has published an interesting set of graphs from some research which shows some correlation between a country’s governance and the reputation of its IP addresses.
The spring ARIN 31 meeting is fast approaching. The final meeting agenda has recently been published. There are also opportunities to participate remotely for those who are unable to make the meeting in person.
Here is my short commentary on the policy proposals being discussed at the meeting. In this blog entry, I’ve also attempted to make some predictions on the discussion and outcome…
Policy Summary: Changes the way utilization is determined for ISPs who return to ARIN for additional IPv6 allocations.
Issues: Since the initial IPv6 policy was implemented, it has been understood that the IPv6 policy would need to be modified as implementation experience was gained. Since the idea of hierarchy is important in IPv6 networks, this policy allows a network’s regions which grow at different speeds to retain the hierarchy structure and still allow fast growing regions to obtain the needed additional IPv6 address space. Since the draft policy’s introduction there was strong consensus that the existing policy needed to change, the challenge has always been the details of policy text.
Prediction: This policy will finally reach consensus at this meeting and will be sent to last-call for approval.
Policy Summary: Allows organizations to request to transfer an autonomous system number (ASN) from one RIR to another.
Issues: ARIN recently adopted policies which both allow the directed transfer of IPv4 between regions and also allowed the directed transfer of ASNs within the ARIN region. This policy extends this trend to allow ASNs to transfer between RIRs. Some stakeholders in general disagree with the idea of allowing IP resources to trade and will likely oppose this policy. On the other side are those who will argue that this policy is a logical extension of the existing policy to allow resources to be transferred to where they are needed.
Prediction: This policy will have signification discussion about the need for the policy and the role of inter-RIR relationships, but I suspect the final consensus at the meeting will be to proceed with the implementation of this draft policy.
Policy Summary: Allows organizations to use a lower utilization requirement for provisioning their 3GPP networks when requesting additional IP addresses.
Issues: Wireless operators have been using space beyond RFC 1918 (such as 184.108.40.206/8) to solve their addressing needs and now that this is becoming part of the “Internet” they need to move off of that space. With ARIN’s /8 inventory currently at approximately 2.5, I’m skeptical that any policy using global IPv4 unicast space can actually solve this problem. The policy text of this draft policy is also not complete at this time and if consensus is achieved on the concepts of the policy change this draft policy would have to return for discussion at another ARIN meeting.
Prediction: This policy will be abandoned by the AC following the meeting.
Policy Summary: Allows ISPs to request smaller than normal IPv6 address blocks or return larger IPv6 blocks to reduce their IPv6 holdings.
Issues: This draft policy addresses an issue where ISPs are being moved into a larger ARIN fee category under the new fee schedule and allows them to return address space to move to a smaller IPv6 fee category. There has been significant discussion on the PPML mailing list about this issue and at this point it seems unclear if this proposal will achieve consensus at the meeting. The primary argument against this policy is that this policy undermines the best current practices for IPv6 subnetting, the intended hierarchical addressing structure defined by the IETF in the IPv6 RFCs, and generous nature of intended IPv6 assignments to end-users. Some stakeholders will argue that ARIN’s fees shouldn’t be used as a force to dictate a network’s IPv6 architecture. Arguments for the policy are that some small ISPs don’t need and never will use the current minimum block size of a /32 or /36 and should be able to get a /48 which meets their network needs.
Prediction: This policy will be sent to last call by the AC following the meeting. (I suspect it is possible that the /48 option will be removed from the draft policy as part of the discussion)
The RIPE region has begun publishing information on the IPv4 transfers which have occurred under the directed transfer policies. This change in published statistics was brought about through the implementation of RIPE policy 2012-5.
A list of the transferred blocks can be found here: RIPE: IPv4 Transfer Statistics
ARIN has posted a set of letters and links in response to a letter from the general counsel of the US National Science Foundation (NSF) that was circulated widely on the Internet in the fall of 2012. This letter was previously written about in a blog entry here and the on Internet Governance Project website.
John Curran, the CEO of ARIN, wrote a letter to the general counsel of the NSF in response to the leaked private letter. In the letter, Mr. Curran requested that the NSF revoke its previous letter or clarify the early IP address assignment context of the letter. The letter also goes on the state ARIN’s case for why it believes it should be the registry of record for these legacy IP resource records and why they should be subject to the same community driven stakeholder policy process as IP address assignments made today by the RIRs.
In a letter, dated November 7th, 2012, the general counsel of the NSF responded to Mr. Curran’s letter and stepped back from some of the statements made previously in the earlier private letter. Specifically noting that the NSF does not speak for the USG on the issue of Internet governance, the NTIA is the appropriate government agency to represent the USG in this area, and that the previous letter was a private letter observation on the NSF’s historical role in the development of the Internet.
This response now seems to erode the idea, that some members of the Internet community have posited, the NSF letter endorsed that legacy IP address assignments should be treated more like property rather than a resource licensed for a specific use.
ARIN’s website on their legacy address information page also now notes the following:
On December 3, 2012, the National Telecommunications and Information Administration (“NTIA”) formally commented on the USG’s Internet protocol numbering principles, including that it recognizes ARIN as the RIR for this region. This NTIA guidance is a clear response to the issues raised by an earlier letter from National Science Foundation General Counsel (NSF GC).
NTIA Administrator Lawrence Strickling posted a blog entry on the US Department of Commerce’s NTIA website clarifying the NTIA’s approach to IP addressing in the US. In the blog entry, Mr. Strickling specifically notes that the USG supports the existing multistakeholder model for development of Internet technical standards and processes, that the RIRs “are responsible for developing policies for the use of IP numbers within their respective specific geographic regions,” ARIN is the RIR for the United States, and that the “USG believes that all IP numbers are allocated for use on a needs basis and should be returned to the numbering pool when no longer needed.”
For those, who still believe that legacy IP addresses should be outside of existing RIR framework and not subject the the needs based policies which have been supported by the Internet community for more than a decade, this series of events only appears to further strengthen the case that the legacy IP address assignments should fall under ARIN’s role as the registry of record and that the USG appears prepared to defend that position in the United States.
Geoff Huston has published his annual look at IP address allocation and assignment statistics.
Plenty of numbers in the report to take a look at… Notably, we saw ARIN’s 2012 (45 million) allocation rate increase back to its 2010 rate after falling dramatically in 2011 (23.5 million). RIPE allocated its last IPv4 blocks under its “regular” allocation scheme in mid-September 2012 and moved into the IPv4 exhaustion phase of allocations. In the RIPE region, there wasn’t an apparent “run-on-the-bank” increase in the allocation rate as the registry moved into the exhaustion phase.
Here Geoff’s updated RIR Address Exhaustion Model shows ARIN moving into the exhaustion phase in mid-2014 with LACNIC in late 2014. AFRINIC’s trend-line currently points to an exhaustion point 9 years from January 2013.
Another interesting statistic found in the report is that the total number of smart phones and tablets purchased during 2012 amounts to almost 779 million units. If each of those devices used a native IPv4 address that would use up 21% of the total IPv4 address space.
Geoff finishes the report with a somewhat pessimistic outlook for the Internet industry.
We are witnessing an industry that is no longer using technical innovation, openness and diversification as its primary means of propulsion. The widespread use of NATs limit the technical substrate of the Internet to a very restricted model of simple client/server interactions using TCP and UDP. The use of NATs force the interactions into client-initiated transactions, and the model of an open network with considerable flexibility in the way in which communications took place is no longer being sustained. Today’s internet is serviced by a far smaller number of very large players, each of whom appear to be assuming a very strong position within their respective markets. The drivers for such larger players tend towards risk aversion, conservatism and increased levels of control across their scope of operation. The same trends of market aggregation are now appearing in content provision, where a small number of content providers are exerting a dominant position across the entire Internet.
This changing makeup of the Internet industry has quite profound implications in terms of network neutrality, the separation of functions of carriage and service provision, investment profiles and expectations of risk and returns on infrastructure investments, and on the openness of the Internet itself.
Iljitsch van Beijnum has recently published an update on IPv6 adoption: IPv6 takes one step forward, IPv4 two steps back in 2012.
The second page of his article mentions the economics of IPv6 deployment as reason for the slow deployment of IPv6.
Apparently, the economics of moving to IPv6 before we absolutely, positively had to without delay weren’t there. As with all technology, IPv6 gets better and cheaper over time. And just like with houses, people prefer waiting rather than buying when prices are dropping. To make matters worse, if you’re the only one adopting IPv6, this buys you very little.
And the pain of the shrinking IPv4 supplies versus the pain of having to upgrade equipment and software varies for different groups of Internet users.
All this means that organizations that are experiencing a lack of sufficient IPv4 addresses will have to address that problem in some other way: by having multiple users share a single IPv4 address through Network Address Translation (NAT).
This echos my thoughts on the matter when I wrote about them extensively in 2011. Use of IPv4 NAT as a substitute for moving to IPv6 was one of the concerns that I and others have raised which could delay or deter IPv6 adoption.
A positive note on IPv6 deployment since 2011: It does seem that IPv6 has been adopted permanently by a number of the large content providers, so at least one side of the transaction is there. Further deployment of IPv6, in my opinion, now depends on the large cable and broadband providers. Only when millions of those subscribers are converted to be IPv6 enabled will we see significant uptake in IPv6 traffic.
Lee Howard of Time Warner noted last year at the IETF 84 that just getting to 1% of current broadband subscribers is a significant effort. At NANOG 56, Lee also noted that the economics of IPv4 xfers vs. carrier-grade NAT vs. IPv6 aren’t as simple as they might seem.
APNIC, in a press-release dated Oct 9th, noted that they had completed their first /24 transfer from ARIN under the new inter-RIR transfer process.
In IGP’s commentary regarding the letter, they claim that this letter confirms a legacy holder’s right to “own their number blocks.” While the letter is certainly an interesting read and its existence is somewhat curious, I don’t really believe this letter resolves anything.
The letter to me clearly states that the NSF general counsel believes that NSI under the NSF contract granted the organization which received the IPv4 Addresses certain rights to use those addresses. Furthermore, the letter goes on to state that NSI did not and could not unilaterally revoke those rights and that the “NSF does not believe ARIN, or for that matter any other organization, could retroactively affect property and rights distributed to you.”
In a comment to the blog post, John Curran CEO of ARIN writes “The concern has never been about ARIN unilaterally reclaiming number resources; it has been about changes to the number resources in the registry and whether such changes must comply with community policy. The letter further does not address in the least ARIN’s operation of the registry…”
My personal opinion is that the ambiguity that does exist regarding the relationship of Legacy IPv4 Address holders who has not signed a registration services agreement with an RIRs will not fully be resolved until a US federal court rules on the specific issues surrounding their status and specific “rights.” It also seems likely that US federal law or regulation could also clear up any ambiguity, but a resolution method through the court system seems more likely. Even then only further litigation can fully resolve any claims a legacy holder or ARIN claim to assert.
I encourage you to read the letter yourself to see what it does or does not say.
The RIPE NCC announced today that they have only one /8 remaining from their available IPv4 pool and have moved to the exhaustion state of allocations. Organizations with additional needs may only request one /22 allocation even if they can justify more IPv4 addresses.
RIPE now joins APNIC who exhausted their free pool in April 2011 as the two RIRs who no longer have available IPv4 for allocation. Recent statistics predict that ARIN should be the next RIR to exhaust their free pool perhaps as early as 2013.
This report discusses what is known about the IPv4 market transfers which have occurred since the implementation of specified transfers within the ARIN, RIPE, and APNIC region. The report brings together a number of sources to come up with the a total of 83 transactions representing 204 blocks for a total of 6 million IPv4 addresses traded between 2009 and the first half of 2012.
The data for IPv4 market transactions is obscured, however, due to the lack of transparency for all transactions. While many have called for additional transparency, full transparency seems unlikely given the current constraints of the RIRs and the lack of other compelling regulations which would require disclosure of IPv4 market transactions.
As expected, the source of most of the transferred blocks are legacy IPv4 holders in the ARIN region. The authors seem surprised that the majority of the transfers occurred in the ARIN region. Given the expected supply for the IPv4 transfer market was expected to come from legacy IPv4 holders (who are largely in the ARIN region) and that an inter-RIR transfer policy was not in place until mid-2012, it is a logical conclusion that the majority of transfers would occur from and to ARIN region entities.
Unfortunately, I believe the study is also laden and interwoven with the authors opinions regarding the market rather than strictly focusing on the facts about the known market.