April 2012 - address-policy-wg

Re: [address-policy-wg] 2011-04 Last Call for Comments (Extension of the Minimum Size for IPv6 Initial Allocation)
by Anfinsen, Ragnar 11 Apr '12

11 Apr '12

+1 MVH/Regards Ragnar Anfinsen > -----Opprinnelig melding----- > Fra: address-policy-wg-bounces(a)ripe.net [mailto:address-policy-wg- > bounces(a)ripe.net] På vegne av Emilio Madaio > Sendt: 4. april 2012 15:22 > Til: policy-announce(a)ripe.net > Kopi: address-policy-wg(a)ripe.net > Emne: [address-policy-wg] 2011-04 Last Call for Comments (Extension of > the Minimum Size for IPv6 Initial Allocation) > > > Dear Colleagues, > > The proposal described in 2011-04 is now at its Concluding Phase. > > You can find the full proposal at: > > http://www.ripe.net/ripe/policies/proposals/2011-04 > > > Please e-mail any final comments about this proposal to > address-policy-wg(a)ripe.net before 2 May 2012. > > > Regards > Emilio Madaio > Policy Development Officer > RIPE NCC >

5 9

Re: [address-policy-wg] 2011-05 New Draft Document Published (Safeguarding future IXPs with IPv4 space)
by Gert Doering 11 Apr '12

11 Apr '12

Dear Address-Policy WG, (cc'ing the EIX wg due to "this is where the proposal came from") after a very lifely debate in the early stages of this proposal, you have been VERY quiet in this review phase. Specifically, *no* comments have been voiced, and this is not enough for the proposal to go anywhere - so unless I see a few more comments really soon now, we'll have to extend the review phase. Please let us know whether this version 3.0 of the proposal is what you want to see become policy, or whether you oppose it. If you oppose it, please make clear whether you oppose the general idea or just specific aspects of the proposal as written now. thanks, Gert Doering, APWG chair On Tue, Mar 06, 2012 at 04:41:23PM +0100, Emilio Madaio wrote: > > Dear Colleagues, > > The text of RIPE Policy Proposal 2011-05, "Safeguarding future IXPs > with IPv4 space", has been revised based on the community feedback > received on the mailing list. We have published the new version > (version 3.0) today. > > Highlights of the changes in version 3.0 are: > > -a new punctuation is used in the first bullet point of the proposed > section 5.6.2 > > > As per RIPE document ripe-500, "Policy Development Process in RIPE", > the suggested change was not considered significant to require a new > Discussion Phase. Hence the proposal moves to the Review Phase of the > RIPE Policy Development Process. > > The draft document for the proposal has been published. The impact > analysis that was conducted for this proposal has also been published > > > You can find the full proposal and impact analysis at: > > http://www.ripe.net/ripe/policies/proposals/2011-05 > > and the draft document at: > > http://www.ripe.net/ripe/policies/proposals/2011-05/draft > > > We encourage you to read the draft document text and send any comments > to address-policy-wg(a)ripe.net before 3 April 2011. > > Regards > > Emilio Madaio > Policy Development Officer > RIPE NCC > > Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279

23 27

2011-05 Last Call for Comments (Safeguarding future IXPs with IPv4 space)
by Emilio Madaio 11 Apr '12

11 Apr '12

Dear Colleagues, The proposal described in 2011-05, "Safeguarding future IXPs with IPv4 space", is now at its Concluding Phase. You can find the full proposal at: https://www.ripe.net/ripe/policies/proposals/2011-05 Please e-mail any final comments about this proposal to address-policy-wg(a)ripe.net before 9 May 2012. Regards Emilio Madaio Policy Development Officer RIPE NCC

1 0

2011-05 Decision to move to Last Call (Safeguarding future IXPs with IPv4 space)
by Gert Doering 10 Apr '12

10 Apr '12

Dear AP WG, 2011-05 (Safeguarding future IXPs with IPv4 space) has come to the end of the Review Phase. Sander and I have gone through the mails sent by you in the discussion phase and review phase, categorizing everything into "support", "opposition", "side track discussion", and so on - as usual :-) - all the details are listed below. Overall, there was very strong support for the proposal, with a fairly high number of supportive e-mails. In the discussion phase, we had a number of people that wanted to see textual changes, or required clarification on details. Some voices had reservation or opposed the proposal initially but changed that to "neutral" or "support" after discussion with the chairs. Two people opposed on the principle of "no special rules" - more to that later. In review phase, 18 individuals expressed their explicit support (some of them relaying support from industry bodies, but this did not influence the decision process). One person sustained his opposition, based on "there should not be a special rule for IXP/ISPs" and "only IXP folks are supporting their support, non-IXPs are underrepresented and would otherwise oppose the proposal". Indeed, many of the voices supporting the proposal have come from people working for IXPs, but I see a number of supporters that I know do *not* work for an IXP - and I also know that there are lots of non-IXP people on the address policy list who are well-capable to express their opposition to a policy proposal. So that part of the argument doesn't hold. Remains the counter-argument of non-supporting special-cases for IXPs - and indeed, the address policy could be much more simple if all consumers of IP addresses were equal. They aren't, and thus we allow special cases if there is enough support for it. Counting voices, I see very strong support. So, we think that we have enough support to call consensus, and move to Last Call. Emilio will do the formal announcement from the PDO soon. If you disagree with our interpretation of what has been said, and the conclusion we have drawn from it, please let us know. Gert Doering, Address Policy WG Chair ----------------------------------------------------- Discussion Phase pro James Blessing Babak Farrokhi Brian Nisbet Remco Van Mook Martin Millnert Sascha Lenz (even if "not a fan of special entities") Niels Bakker Tom Hodgson "worried about special cases, but still supports policy" Henk Steenman Wolfgang Tremmel Larisa Yurkina Harald Michl David Freedman Job Snijders Vincent PAGES Barry O'Donovan Michiel Appelman Steven Bakker Tom Bird Erik Bais Jan Zorz Carlos Friacas David Croft Basile Bluntschli Rolf Schärer Daniel Holme Kjetil Otter Olsen Sascha Luck Nick Hilliard (with some reservactions about the "must do IPv6" bits) against Florian Weimer "IXPs shouldn't be special-cased" chrish(a)consol.net "no special treatment... 'no'" "this should not be necessary" Nigel Titley (later agrees with intent of proposal) Jim Reid ("reservations but no strong objection") clarifications Dan Luedtke ("which /16 is used?"), answered by proposer Remco Van Mook (sizing criteria?), answered by proposer and chair, text integrated into next version Martin Millnert (what about multiple peering LANs?), answered by proposer Sascha Lenz (renumbering time?), answered by Niels Bakker Nigel Titley ("shouldn't be the every-lir-gets-a-/22 rule be good enough?"), +1'ed by Jim Reid, answered by proposer Henk Uijterwaal ("will v4 IXPs be needed after v4 runout?") side-track "why does an IXP have to return PI and not PA?" Florian Weimer, Niels Bakker side-track "shouldn't this be limited to non-commercial IXPs?" Erik Bais, Andy Davidson, Jim Reid side-track "address assignment must be fair, no special rules for IXPs, and non-IXP-folks are not properly represented at APWG anyway" chrish(a)consol.net, Nick Hilliard (multiple mails) Review Phase, v3.0 with textual changes suggested in discussion phase pro James Blessing Sascha Luck Jan Zorz Sascha Lenz Remco van Mook Florian Hibler (needs more coffee) Paul Thornton Arnold Nipper Job Snijders Alexander Leefmann Lindqvist Kurt Erik Christian Kaufmann Kjetil Otter Olsen Petr Jiran Jérôme Fleury Jan Torreele John Souter Sylwester Biernacki against chrish(a)consol.net side remark Turchanyi Geza "it's good that the WG affected is supporting the proposal" (side track discussion about organizations or individuals speaking in the APWG) -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279

1 0

On use of the word "sold" (was: 2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers))
by John Curran 07 Apr '12

07 Apr '12

RIPE Policy Folks - I have no comments to make with regards to the merits of policy proposal 2012-01, but will comment for sake of clarity on the use of the words "sell" and "sold" in the proposal. In the policies in the ARIN region, we have made use of the term "transfer" for this purpose, although one can argue that "sell" and "sold" is equally accurate. The challenge with using the word "sell" is simply that for many readers it carries the connotation of 'free and clear' transfer of property to another. I do not know, (nor intend to assert) what transfer of number resources in the RIPE region means, but in the ARIN region we hold that there are multiple parties with rights to each number resource, and this includes the resource holder (who has the right of exclusive use) but also includes the community (who has the right of visibility into the public portion of the registration, the right to set policies by which the resource may be transferred or reclaimed, etc.) We've had parties come to ARIN and indicate that they have "bought" or "sold" their address blocks, and we have to point out that they may indeed have sold their particular rights to the address block, but that any transfer in our region must comply with the community policies, and ask that they put a transfer request (agreed to by the seller) so that we may transfer the number resources to the new registrant. In short, using terms like "sell" and "sold" in the policy text may create the impression for some readers that Internet number resources are just widgets to be bought and sold as desired, but that error is may be unavoidable in any case given the circumstances. I do not believe that the use of "sell" or "sold" is incorrect, but simply of question of personal judgement as to whether the improved readability is worth the connotations that some readers will draw as a result. FYI, /John John Curran President and CEO ARIN

6 9

2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers)
by sandrabrown＠ipv4marketgroup.com 06 Apr '12

06 Apr '12

2 1

2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers)
by sandrabrown＠ipv4marketgroup.com 05 Apr '12

05 Apr '12

Sorry, in plain text this time.... Thank you for your comments on proposal 2012-01, which I have submitted. I will undertake to speak to several of the points you have raised. My opinion on the question of the use of the word "sold" is that I would also be ok with the word "transfer". I believe this policy can cover two situations: 1) where related companies, for example ABC-NA transfers a block from ARIN to ABC-EMEA to RIPE-NCC, and no money is exchanged; and 2) where unrelated organizations transfer a block. Only in the second case is money likely to be exchanged in order to encourage the sending organization to initiate the transfer. Thus only the second case is truly a sale. And let's be very clear that I do mean for money to be exchanged: in most cases there is work to be done to free up blocks of IPv4 addresses for re-allocation, and the incentive for entities to do so is financial benefit. Otherwise, large portions of their blocks will continue to sit dormant and unused. The entire purpose of the IPv4 address market is to free up dormant, unused IPv4 addresses to fill needs. A number of people have questioned the 15 month window before allowing a buyer to re-sell. I have had suggestions ranging from 12 months to 36 months. I think the point is that there is some feeling that an entity could acquire IPv4 addresses with the intent to flip them for profit. I am not sure this would be possible since at some point, the transition to IPv6 will make IPv4 addresses a declining value asset. I, for one, believe in a free market for IPv4 addresses, where no holding period is required. In a free market for IP's, the price of IP's would move up and down like the price of a commodity. However, there are artificial trading rules at play with IPv4, so some window seems to be needed. Hence, I chose 15 months as a compromise for those who feel that 15 months allows a company to satisfy its IPv6 conversion needs and possibly be done with IPv4 and be ready to sell them. To Mr. Blessing's question about an LIR changing its policy, I believe the answer is that there would be a legal binding agreement between the buyer and the seller that would cover this situation. The RIPE-NCC policy does not need to cover the legalities between the two parties. I am not sure what the question of a reverse policy refers to? Do you mean to reverse the transaction? If so, you could add a clause that states that the two parties can reverse the transfer at any time in the future or something to that effect. I did have a hypothetical discussion with Sanjaya of APNIC for a situation that may require this. I was asking about a potential lease situation where an APNIC member may wish to lease a block from another RIR. He wondered if the APNIC registry could initially be updated to reflect the lease, and then once the APNIC member had completed the upgrade to IPv6, the IPv4 addresses could be returned to the originator. We then wondered how to reverse the transfer to return the IP's to the owner. This clause above that Mr. Blessing is possibly questioning about would take care of lease returns. I believe these transfers would always be between entities, ie. LIR's. Mr. Hilliard is concerned about Afrinic losing IP's. Frankly, I have not looked at oversupply or demand in the Afrinic region. Right now Afrinic does not have an Inter-RIR transfer policy and by including a statement that RIPE-NCC will only transfer with RIR's that have a policy, you allow Afrinic to take care of itself. Mr. Lenz is concerned about the use of the word "business" purposes. That is, perhaps, a semantic point. I believe that the word business should not be taken too narrowly. Business may be defined as an activity a person or organization is engaged in, and I would just take it at that. Mr. Leefman suggested we realign the proposal to a /21 as the minimum size, and I think that is an excellent suggestion. Other RIR's such as ARIN and APNIC work with a minimum size of a /24 and I was trying to keep RIPE in line with them, but frankly that is such as small block, why bother? Mr. Leefman also suggested changing the requirements for proving need to align with the intra-RIR policy. In checking with Mr. Madaio, he informs me that there have been no intra-RIR transfers to date. That suggests to me that RIPE's intra-RIR policy needs some rework, if I may be blunt. My experience with transfer markets in other RIR service areas has shown three things: 1) that entities need to be assured that their transfers will be approved, when they put money on the line, 2) that when they pay, they don't want, or believe that they should have to, go through the same needs justification process as when they apply for free allocations, 3) while the period of a free allocation may be fixed (ie 3 months), when a business has a need and a business plan for which it acquires IP's it is a much longer time horizon, and it is not RIPE-NCC's role to limit its members business planning horizon to 3 months. (If RIPE approves me for a three month supply now, I have no guarantee I will get another 3 month supply in 3 months....I cannot plan or operate this way!). For example, and you may not care about what other RIR's do, but ARIN has a 24 month supply for a transfer, and as you may know, we are working to convince ARIN to modify its policy on needs justification for transfers. So, I would not support a change in the inter-RIR policy to match your intra-RIR policy. I would instead suggest that you consider modification of your intra-RIR policy to match what your constituents need. Lastly, I like Mr. Van Mook's comments, other than that he opposes the policy. It is excellent and simple to outline the transfer in and transfer out options, and as noted above, I concur with him that the RIPE intra-RIR policy needs clean-up. I guess I thought that was a bigger mountain than I could climb, but if it is possible to have a conversation about doing it at the same time, I am all ears. Again, thank you for your support, your comments, and your thoughtful feedback. Respectfully, Sandra Brown

1 0

Re: [address-policy-wg] 2011-04 Last Call for Comments (Extension of the Minimum Size for IPv6 Initial Allocation)
by Sascha Lenz 05 Apr '12

05 Apr '12

Hi all, > > Dear Colleagues, > > The proposal described in 2011-04 is now at its Concluding Phase. > > You can find the full proposal at: > > http://www.ripe.net/ripe/policies/proposals/2011-04 > I didn't have a change of mind lately, still happy to support it in this form. (Yes, i'm selfish, "more bits to brag with^W^Wmake deployment easier in some cases", no downsides - /29 is reserved anyways for existing LIRs) -- Mit freundlichen Grüßen / Kind Regards Sascha Lenz [SLZ-RIPE] Senior System- & Network Architect

1 0

Re: [address-policy-wg] 2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers)
by Remco Van Mook 04 Apr '12

04 Apr '12

While I would strongly support the development of an inter-RIR transfer policy (and/or cleanup of the current intra-RIR transfer policy), I don't think this proposal is sufficiently clear and in line with other current policy in the region to work. In other words, I oppose this proposal. (just to be clear for the WG-chairs who are keeping count) What I would like to see instead is a simple framework, outlining the two options: 1) transferring TO the RIPE NCC service region 2) transferring FROM the RIPE NCC service region that clearly define which parts of the regular transfer policy would apply to such a transaction. With that, we keep all the rules about transfers in one place. 1) would be something like 'We welcome any address space transferred to the RIPE NCC service region as long as transferring it is in line with policies at the originating region and the entity receiving the space qualifies as outlined in the policy for address transfers within the service region' 2) would be something like 'You can transfer any address space to outside the RIPE NCC service region as long as transferring it is in line with policies at the receiving region and the address space qualifies for transfer as outlined in the policy for address transfers within the service region' Finally,if you then feel unhappy with the limitations in the current transfer policy, I would very much like to invite anyone to introduce a policy proposal to remove them. Best, Remco van Mook Director of Interconnection, EMEA remco.vanmook(a)eu.equinix.com +31 61 135 6365 MOB EQUINIX 51-53 Great Marlborough Street London, W1F 7JT, United Kingdom On 02-04-12 15:39, "Marco Schmidt" <mschmidt(a)ripe.net> wrote: > >Dear Colleagues, > >A new RIPE Policy Proposal has been made and is now available for >discussion. > >You can find the full proposal at: > > http://www.ripe.net/ripe/policies/proposals/2012-01/ > >We encourage you to review this proposal and send your comments to ><address-policy-wg(a)ripe.net> before 30 April 2012. > >Regards > >Marco Schmidt >on behalf of the Policy Development Office >RIPE NCC > > This email is from Equinix Europe Limited or one of its associated/subsidiary companies. This email, and any files transmitted with it, contains information which is confidential, may be legally privileged and is solely for the use of the intended recipient. If you have received this email in error, please notify the sender and delete this email immediately. Equinix Europe Limited. Registered Office: Quadrant House, 4 Thomas More Square, London E1W 1YW. Registered in England and Wales, No. 6293383.

2 1

[apwg-chairs] 2011-04 Decision to move to Last Call (Extension of the Minimum Size for IPv6 Initial Allocation)
by Gert Doering 04 Apr '12

04 Apr '12

Dear AP WG, Sander and I have looked at the mails sent by you about the proposal 2011-04 (Extension of the Minimum Size for IPv6 Initial Allocation), and I tried to categorize everything hat has been said into bins like "support", "opposition", "side track discussion", "clarification" - the result of that is at the end of this mail. In the initial discussion phase, I counted nearly 150 e-mails. Of those that expressed a clear "support" or "opposition" opinion, the majority were in support for the proposal, thus we moved on to review phase, with some textual changes as suggested by you. In review phase, 3 people voiced their support, nobody voiced objections, and none of the people that had objections in discussion phase stated their opinion (neither positive nor negative). Given that, we assume that those that had reservations agreed to disagree but didn't want to stop the proposal, or agreed with our assessment that we have enough support to go ahead. The main argument for objection that we could make out was "this is not good stewardship, to give people more addresses than they need, without providing any documentation". I can see the argument, but I'm also convinced that part of good stewardship is 'finding the right balance between "too big" and "too small"' - and since a number of operators have spoken up and said that being able to increase their /32 to a /29 "just so" would ease their deployment, we took that as a strong enough counterargument. The argument voiced that "we should not make a specific proposal for 6rd" was already taken into account - the proposal is technology-agnostic. So, we think that we have enough support and no sustained opposition to call consensus, and move to Last Call. Emilio will do the formal announcement from the PDO tomorrow. If you disagree with our interpretation of what has been said, and the conclusion we have drawn from it, please let us know. Gert Doering, Address Policy WG Chair ----------------------------------------------------- Discussion Phase pro Michael Adams (want bits, not 6rd) Roger Jørgensen Anfinsen, Ragnar Sascha Lenz (want bits, not 6rd) Andy Davidson (with "ask for it" clause) Daniel Holme (no protocol specifics!) James Blessing (with additions for existing address holders, taken into v2.0 Jasper Jans David Freedman Roman Sokolov Tero Toikkanen Job Snijders Tom Hodgson (more bits, not tied to mechanisms) Wilhelm Boeddinghaus Rinse Kloek Sascha Luck Florian Fuessl Jerzy Pawlus Daniel Roesen against, with "want to see proper documentation that says 6rd" James Blessing [answered by Jan Zorz, discussion ended, later agreed] Martin Millnert Mikael Abrahamsson [give everybody who says "6rd" a /31] Nick Hilliard [answered by Jan Zorz, "is ask-for-it enough?"] Tim Chown against, with "make it easy to get a /29, but not automatic, ask for it" Remco Van Mook [answered by Jan Zorz, taken into v2.0 -> agreed] against, with "6rd is a bad protocol and there should be a temporary block" Turchanyi Geza [answered by Jan Zorz, "this is not 6rd-specific", Ole Troan, Ragnar Anfinsen] clarification Mikael Abrahamsson (reservation of a /26?) [answered by chair and proposer] Andy Davidson (would reservation be OK from global policy PoV?) Martin Stanislav (please clarify exact rules for /29../31, answered by prop.] unclear whether in favour or against Dan Luedtke Martin Millnert (*if* 6rd as a special case, make it temporary-only) [answered by Jan Zorz] Daniel Suchy chrish(a)consol.net "not big enough" side-track Randy Bush (let's go to a /16) [answered by chair, David Conrad, Roger Jørgensen, Jan Zorz, Florian Weimer] "make subsequent allocation policy easier (instead of initial?)" side-track Dan Luedtke [answered by Michael Adams, chair] Daniel Roesen [answered by chair, Jan Zorz, Remco Van Mook] "does 6rd need a /31 or /24?" side-track Mikael Abrahamsson, Jan Zorz "which operators are asking for 6rd" side-track Martin Millnert, Jan Zorz, Michael Adams "alternatives to 6rd" and "is it infinite?" side-track Ahmed Abu-Abed [answered by Jan Zorz, Remco Van Mook, Ole Troan, Nick Hilliard, Roger Jørgensen] "why not make it a /26 or /24 right away?" side-track Roger Jørgensen, Leo Vegoda, Nick Hilliard, Daniel Roesen, Jan Zorz ------------------------------------------------------------------- review phase pro Michael Adams Tiberiu Ungureanu Gleb Katenin side-track: "subsequent allocations Michael Adams, Tiberiu Ungureanu, Daniel Roesen, Jan Zorz, Nick Hilliard side-track: "how much address space does google have?" chrish(a)consol.net, Gleb Katenin -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279

3 2