______________________________________________________________________ DRAFT TRANSCRIPT APNIC Member Meeting Friday 9 September 2005 2.00pm ______________________________________________________________________ PAUL WILSON: Good afternoon, and welcome back. We will continue now with the afternoon agenda. We've got one item from this morning, which is, of course, the Address Council report by Kenny and we'll be going straight into the SIG and BoF reports after that. So Kenny, the Address Council report, thank you. KENNY HUANG: Thank you, Paul. Good afternoon and my name is Kenny Huang. I'm going to give a report on the Address Council. And there are three representatives and the other is Takashi Arano and the other is Hyun-joon Kwon. So the ASO Address Council does not represent the RIR directly. There are three representatives from each RIR region and, after the signature of the new MoU, which was signed on 21st of October in the year 2004, a new organisation will now have two elected and one nominated by each RIR board. So the representatives are not necessarily RIR members and all interested parties may participate. ASO Address Council is responsible for the organisational roles of providing recommendations to the board of ICANN concerning the recognition of new RIRs, according to the agreed requirements and policies as currently described in the document ICP-2; and defining procedures for the selection of individuals to serve on other ICANN bodies, in particular on the ICANN Board and implementing any roles assigned to the Address Council in such procedures; and providing advice to the Board of ICANN on number resource allocation policy in conjunction with the RIRs. And the last one is developing procedures for conducting business in support of their responsibilities, in particular for the appointment of an Address Council chair and definition of the chair's responsibilities. All such procedures should be submitted to the Executive Council of the NRO for approval. NRO formed by five RIRs - Regional Internet Registries - to formalise their cooperation efforts - "the RIRs in concert" - the NRO exists to protect the unallocated number resource pool and promote and protect the bottom-up policy development process and act a focal point for Internet community input into the RIR systems. Policy for allocation of IPv4 blocks to RIRs - the policy describes the procedure by which IANA will make IPv4 allocations to the Regional Internet Registries. The draft policy for IPv4 was submitted by the NRO - Number Resource Organization - to the ASO Address Council. And ASO resolution has been approved by ASO in accordance with the four RIRs' local policy development processes - sorry, it should be five RIRs' local policy development processes. Global addressing policy was forwarded to ICANN Board of Directors on the 3rd of September in the year 2004. ICANN Board of Directors final resolution in Mar del Plata on 8 August, in the year 2005. For IPv6 global policy, ASO community is considering global IANA IPv6 address allocation policy and, through the Regional Internet Registry public policy forums, the ASO community is considering global policy proposals that will govern how IANA allocates IPv6 address blocks to the RIRs. Proposals have been introduced in all of the RIR communities. The proposals are currently being discussed through RIR mailing lists and the global IPv6 mailing list. ASO and NRO encourage all who are interested in IPv6 policy to join these discussions and to participate in the RIRs' public meetings where this issue has been added to the agenda. So AfriNIC - ASO and NRO a pleased to note AfriNIC's continuing progress since it received ICANN recognition during the meeting in Mar del Plata, Argentina, as the fifth Regional Internet Registry at April 8, year 2005. AfriNIC II was held at 26 and 27 of April 2005 in Maputo, Mozambique. The AC is pleased to welcome the elected AfriNIC AC members Jean Robert Hountomey for three-year term, Sylvia Geha Kezengwa for two-year term and Alan Barrett for one-year term. So that's the list for the current members of the Address Council and it's quite a long list so we can check it on the website (refers to slide). So ASO administration through the Secretariat. The Secretariat for year 2005 is supported by LACNIC. And this is rotated between the RIRs. For example, in 2002, the Secretariat was ARIN. And APNIC for year 2003. RIPE in the year 2004 and LACNIC in 2005. AC elected ICANN board members including Raimundo Beca from LACNIC and Mouhamet Diop from Africa. The meeting of the AC - we have bi-monthly telephone conference and we have yearly two workshops and, like, 27 April in Maputo and in November it will be held a workshop in Vancouver. For outreach and community input - RIR policy meetings and ICANN meetings. So recent issues from the Address Council regarding Ð responds to the ICANN strategic plan document. So first draft comments are prepared by Sebastian Bellagamba and discussion by AC Address Council using e-mail before the final publication. And also focus on the WGIG -Working Group on Internet Governance - and advice to the ICANN Board of Directors on addressing aspects of the WGIG work and also appointed Lee Howard to attend the WGIG meeting on behalf of the AC. So input to ICANN Board and comments to ICANN procedure to handle ASO policy proposal - ICANN posted a review procedure and Louis Lee and Sebastian drafted an AC response to the procedure and we also discussed at the AC workshop in Maputo and formally approved at regular AC conference call at May 4th, year 2005. So ongoing Address Council issues include advice to the ICANN Board on a new gTLD strategy implementation, so we do analysis of IP addressing issues or concerns in the proposal and also advice to the ICANN Board on competitive IPv6 allocation and assignment policies. We have formed a working group to study on these issues. And follow the development of IPv6 global policy. We have presentation at all RIR meetings. The ASO Address Council shall review the process followed by the RIRs in terms of reaching a position of common agreement and common text to describe the proposed global policy and the Address Council shall undertake measures in accordance with adopted procedure to assure itself that the significant viewpoints of interested parties were adequately considered, in other words document in MoU. OK, that's the end of my AC report. Any questions? PAUL WILSON: OK. Thanks, Kenny. The next session is - the next section of the agenda is the SIG and BoF reports and Kenny is the chair of the policy SIG. So you can continue. Thanks. KENNY HUANG: OK. Here we go to the 12th policy SIG report. And myself, Kenny Huang, and Toshi-san is the co-chair of the policy SIG. That's the history of the whole policy SIG from first one in Korea and we are in 12th That's the agenda on the policy SIG (refers to slide) I'll quickly go through that Ð Review of open action items. And IANA policy for allocation of IPv6 blocks to RIRs. And proposal to amend APNIC IPv6 assignment and utilisation requirement policy. And alternative address allocation algorithm for IPv6. Resource recovery policy - implementation update. And second session Ð Study the proposal from discrete networks and national peering. RIR policy development update. LIR survey results. IP address space transfer between existing Taiwan LIRs. And large IPv4 address space usage trying for future IPv6. And the last one would be the election of the co-chair. So that's the head count for each session. We had two sessions. In the first session, the head count was 66. In the secretary session, it's 86. That's the content of the SIG - we had three policy proposals - it actually became four because one of the proposals was split into two. So we had four policy proposals. We have six informational proposals. Two of the policy proposals reached consensus in the OPM so we need to look for consensus at the AMM. Another two policy proposals didn't reach consensus in the OPM. OK, that's the first policy proposal. It's IANA policy for allocation of IPv6 blocks to RIRs - and that's presented by Paul Wilson. Actually, this proposal, we had been discussing in a previous SIG meeting and it went through the different RIR regions to circulate and discuss the proposal. So this would be the revision one. The proposal itself - allocation sufficient for 18-month period instead of 36 months and those with less than /12 unallocated address space to receive an allocation from IANA. RIRs to receive additional IPv6 space when available space is less than 50% of a /12 or available space of IPv6 address is less than RIR's requirement for the following nine months. So that's the revision one. The situation in the other RIRs including AfriNIC - that will be discussed in the next AfriNIC meeting. For ARIN, the proposal already reached consensus and for LACNIC it already reached consensus. And RIPE is still having discussion in RIPE 49, 50 and the coming 51. So this proposal reached consensus at the policy SIG OPM. The second policy proposal was proposed by Geoff Huston - a proposal to amend APNIC IPv6 assignment and utilisation requirement policy. And the proposal itself was split into two proposals so I need to go through these two proposals. Proposal 1 was evaluation for subsequent allocations to be based on an HD-ratio value of 0.94. So the first part of the proposal reached consensus at the OPM. The second part of the proposal - adding a /56 end site allocation point and default end site allocation for SOHO end sites to be /56. This second part of the proposal didn't reach consensus at the OPM. We had another policy proposal for discrete networks and national peering proposed by Uchenna. The proposal "recommended that APNIC allocate routable aggregate blocks of IPv6 to enable global/continental-regional national peering for large ISPs with global facilities" and the proposal didn't reach consensus. So we had six informational presentations so just to quickly summarise - the first one was "alternative address allocation algorithm for IPv6" presented by Mei Wang. The second was resource recovery policy implementation update by Son Tran and the third was RIR policy development update presented by Save. The next one - LIR survey result reported by Save and we also have action item - continue to discuss on the mailing list. The next one is IP address space transfer between Taiwan LIRs presented by Ching-Heng Ku from TWNIC. And the last one is large IPv4 address space usage trial for future IPv6 presented by Naota-san. The last was election of co-chair. We had four nominees and eventually the elected co-chair was Eugene Li. So Eugene Li will serve in the next policy SIG meeting. So we also have three action items - pol-20-001 - "Pending approval at each remaining stage of the policy proposal process, Secretariat to continue coordinating global acceptance of this policy - prop-005-v005." And that's Paul Wilson's proposal. Pol-20-v002 "pending approval at each remaining stage of the policy proposal process, Secretariat to continue coordinating global acceptance of the HD-ratio component of this policy and that's prop-031-v 001 by Geoff Huston and poll-20-002 "APNIC Secretariat to refer further discussion of the LIR survey to the mailing list, that's prop-020-v001. So we come it the final one, the next step. The policy proposal get consensus from the OPM. We need to seek another consensus from the AMM. We have two policy proposals. The first one is prop-005-v005 - IANA policy for allocation of IPv6 blocks to the RIRs and that's proposed by Paul Wilson. And the second part is prop-31-v 001 only part one - modifying IPv6 HD-ratio from 0.8 to 0.94 and that's proposed by Geoff Huston. So we need to seek consensus from the AMM. PAUL WILSON: Thanks, Kenny. To explain this process to those who may not be fully familiar with it - the role - one of the roles of the Member Meeting is to review the consensus outcomes of the policy SIGs and to decide whether it - the Member Meeting - will also agree by consensus with those outcomes from the SIGs. Even if the agreement is reached here, there is a further period of four weeks of comment on the mailing lists for the SIGs concerned and after that a review of the entire process by the APNIC EC, who then endorse the outcome of the process providing that there has been no problem with the process and providing that the consensus is seen to be common across all of those areas. It's possible, for instance, that consensus might be reached here and then there is some substantial observation raised on the mailing list, in which case the APNIC EC may possibly decide that this policy SIG - the policy outcome should not be endorsed and should go back to the meeting. OK, so we're at the stage now of reviewing the consensus outcomes of the SIGs and that's a process whereby I will ask this meeting for a show of hands in agreement and then in opposition to each of the policy proposals. The first one to look at is prop-005-v005 - the IANA policy for allocation of IPv6 blocks to RIRs. So could I ask for an indication from this meeting of agreement with this - for those who agree with this policy proposal, please raise your hands. The APNIC staff are just counting. OK, thank you. And, if anyone disagrees with the policy proposal, please raise your hand. OK, I think we have consensus. Thank you. The second policy proposal from the policy SIG is proposal 31 version 1 and we're looking at only part one of the original proposal. The Address Policy SIG came to consensus on part one which was to change the IPv6 HD-ratio from 0.8 to 0.94. So could I ask for indications of support for this proposal. If you support the proposal, would you please raise your hands. Once again, they're all being counted by the APNIC staff who are at the side there. OK, thank you. If you disagree with the proposal, could you raise your hand. Well, it seems to me that the numbers voting in this case are not huge. There were certainly more hands in favour than against, substantially more. Would interpret that as a consensus decision in this case. However, as I have mentioned, this is not the end of the story. So a consensus of this meeting is then subject to further discussion and comment on the mailing list and then a review from the APNIC EC. Could I ask if there's any comment - if anyone would like to comment, make a suggestion, observation, about this process or this particular policy. KAZU YAMAMOTO: This proposal originally came with /56 issue and my question is whether or not we can discuss this stuff separately from /56. For example, if the HD-ratio is calculated based on /48 but I'm not sure I'm affirmative to 0.94 which is based on /56. I think we cannot - I think it is hard to discuss this stuff separated from /56 stuff. So my suggestion is to discuss both together on the mailing list. That is my point. PAUL WILSON: OK, thank you. GEOFF HUSTON: A point of clarification here - during the address policy SIG, 0.94 HD-ratio was put to the policy SIG without any reference to a 48 or a /56. The reference was to end sites. So it was not tied to a /56. When that was put to the SIG policy meeting. RANDY BUSH: Randy Bush. If you need a number to calculate the HD-ratio, - if you need some prefix length to calculate the HD-ratio, why not use 42 - it's traditional. GEOFF HUSTON: Geoff Huston again. Again I reiterate, there was no reference to what the end site size was. What we were proposing and what I believe the question put to the policy SIG was - HD-ratio of 0.94, which is the density of end sites. Was it. There was no specification of the size in what was put to the SIG policy in the meeting. KENNY HUANG: And that got consensus from the OPM. PAUL WILSON: OK, well, as suggested, there'll be further discussion on the mailing list. It is difficult, I think, when a policy proposal comes along that has got two parts and it, on the fly, the two parts are considered separately, which assumes that they are independent and can be considered independently. I'd say we have consensus on the - we have a rough sort of consensus on part one here but it's subject to discussion on the mailing list so that's what I suggest we do, we continue in that matter and, after that four-week period, it's up to the APNIC EC to decide whether or not the consensus reached at the meeting should be endorsed after the discussion. The SIG itself continues on the mailing list and it may well - the consensus position may well change or be maintained. OK, thank you. Thank you very much, Kenny and, before you leave, I have a small gift, as is traditional. Thank you. APPLAUSE The next SIG on this list is the database SIG so, if you're ready Xing... And perhaps other SIG chairs, if you could just note the list on the screen here in terms of what's coming up. George will be on after Xing. XING LI: OK, good afternoon, this is the report for the Database SIG. This year, the job was quite easy. The remaining action items before this SIG meeting is db-19-001 - "following approval at each remaining step of the policy development process, the APNIC Secretariat will implement proposal - APNIC to publish address assignment statistics." And this is done. And the second one is database 18-001 for the IPv6 Internet routing registry and this is also done in the testing phase. Another action item is database 17-002 - the mirroring of the Internet routing registry and the status is this open item withdrawn by the proposer. In summary - the proposal 025-v001 - the IPv6 routing registry, the test Whois database can be found in testwhois.apnic.net. And to summarise the form of address assignment statistics and, if you are interested for the research progress, you can download from ftp.apnic.net. So, that's it. Thank you very much. Oh, one remark - it seems Database getting - the load is getting lower and lower so I don't know if there are some other comments or suggestions. Thank you very much. PAUL WILSON: As most of us would have seen, there is a call for proposals for SIG presentations, which is issued some weeks or months before this meeting actually takes place. So, in this case, we didn't have too much content provided or contributed to the database SIG. If any of you have some database SIG issues, proposals, reports, then please - I would encourage you to respond to the next call for proposals and come forward and make a contribution to the database SIG you would be very welcome. Thanks. George is coming up to report on the DNS Operations SIG. That will be followed by the IPv6 Technical SIG. GEORGE MICHAELSON: Hello, everyone. I'm giving this report on behalf of Joe Abley who wasn't able to be present for this meeting. We had a very good meeting. There were about 50 participants. We had three substantive matters that got discussed. There was proposition 30, which is about the deprecation of the ip6.int reverse DNS service in APNIC. This was presented by Sanjaya and, if it was adopted we had some outcomes on APNIC to work with the fellow RIR towards this main suitable point and also regarding statistics and to try and contact domainholders. There was good consensus to proceed with this from the participants in the SIG, with report back at AMM 21 and a recommendation that we should investigate the current utilisation of these domains, spend more effort to look into how people were looking up these reverse domains, where they came from and find out if we could get in touch with them. I delivered a report on the lame DNS policy, showing that we've managed to achieve a reduction of lame DNS in the region from 16% to 8% and an outcome is a request that we try to gather some more statistics on the address space, particularly the domains that have never been delegated. Mathias Koerber presented a discussion on issues around scaling the DNS and has requested that people with an interest in future DNS developments and deployment of services based on the DNS be in touch with him and discuss the requirements for the whole community to try and look at the developments that would improve DNS service worldwide. So, given that we had strong consensus in this SIG, we have proposition 30 for the deprecation of the ip6.int reverse DNS service in APNIC. This was proposed by Sanjaya. We had a good consensus in the SIG for adoption of this activity and if we adopted that here that would be two outcomes. There would be an ongoing implementation process subject to approval which would be implemented by the Secretariat and there would be statistics and reporting process that would come back at the DNS SIG at APNIC 21. PAUL WILSON: OK, thank you, George. So we do have one consensus outcome from this SIG, which is the deprecation of the ip6.int domain so could I please ask for hands in favour of this proposal. Thanks, OK. Could I ask for hands against this proposal please. OK, thank you. I think that's a strong consensus support for the proposal. I mentioned in my report earlier Anne Lord, who is not - wasn't able to be with us in the meeting today but she is with us virtually. She has corrected me that the period of comment on the mailing list is eight weeks, not four weeks so, in case of any of the three consensus policy outcomes that you've seen, we've got eight weeks for a comment period on the mailing list. Thank you, Anne. And thank you, George, for your contribution. (Offers George a gift) GEORGE MICHAELSON: They don't go to staff. Great. Thank you. APPLAUSE PAUL WILSON: While Kazu is setting up, I'll just do what I should have done before - I clearly need a bit more help up here, but this is a gift for Xing as well. APPLAUSE KAZU YAMAMOTO: So, IPv6 Technical SIG, which was held on Wednesday morning. Well, the number of audience is about 50. That is very good. And we have no action items. And we have five presentations. So chair and co-chair thank speakers. Thank you. Let me explain briefly some of them. First one is IPv6 allocation status report. This is a semi-annual report from APNIC. This time, Sunny made a presentation. And his presentation covers some statistics and routing table issues and Whois registrations. This is very good survey on IPv6 so if you are interested in it, please give a look at his presentation material on the website. And the next presentation came from Sawabe-san from the IPv6 Promotion Council of Japan. His presentation covers IPv6-ready products in Japan, including web camera, remote controller. And IPv6 showroom provided by promotion council. And IPv6-related activities, including voice interoperability. And a distributed weather information system. OK, then the next presentation was given by Jordi. He talked about IPv6 distributed security. He pointed out some drawbacks of centralised security. Currently we are using. So to solve this problem, a European Commission made a project to solve this kind of problem and he explained several host-based security models. OK, last one is provided by Hosaka-san from JPNIC. His title was ÔJPNIC IPv6 registry service experienceÕ. Well, he explained JPNIC's agent service, which was started in 2000. This service is including a reverse DNS service and a Whois registration. Well, advantage to Japanese people is registration can be done in Japanese. This is very, very big to Japanese people. And well, after this service started, a massive growth - we can show massive growth in registrations with this service. So to my impression, if another NIR people have this kind of service to each country, that is very good. So, that's it. PAUL WILSON: OK, thank you, Kazu. APPLAUSE Now Izumi Okutani will present the NIR SIG, followed by three reports from Philip. IZUMI OKUTANI: Hi. I'm Izumi Okutani and I will be presenting what we discussed at NIR SIG on Tuesday this week. It was generally a well-attended session with roughly 30 attendees and this time we had quite a few attendees who are not NIR members, which was a pleasant sign to find. And a total of four presentations were made during the session, one of which was a proposal. So we had a consensus on this particular proposal, which was to totally abolish IPv6 per-address fee for NIRs just until it is necessary. The other thing to note is that actually we also had four other presentations planned for this session, but we ran out of time. So we really need much more time for more presentations in the coming meeting. So now I would like to explain a little bit more about what we discussed. This is a list of the presentations made. And, as you can see from the titles, the discussions were mainly focused on the revision of NIR fee. The first presentation was an informational presentation made by Paul Wilson from APNIC and he explained the recent fee revision that was implemented by APNIC based on the EC decision in April this year. This was to completely abolish what is called per-address fee for confederation members and the second point on - and the second decision made was that apply 90% discount for per-address fee that is charged for IPv6 allocations which is based on IPv4 infrastructure. And I'm sure to explain all this doesn't really make sense to some of you who are not familiar with NIR fee structure so I will explain a little bit more in detail about how the fee structure for NIR works. So please bear with me a little bit more on this explanation. And there was no questions or comments from the audience on this presentation. The second was the proposal made which was jointly proposed by NIRs and although the 90% discount was applied recently by APNIC, NIRs felt that this was not sufficient to address the problem. They feel they are facing. Of the reasons was that this recent 90% discount scheme is too complicated and they wish to completely abolish the per-address fee for IPv6 instead of this applying a discount. OK, I will explain this separately in details in later slides. And third, the NIRs feel we need a long-term revision for the fee structure for NIRs instead of trying to apply short-term solutions so, again, Paul Wilson from APNIC presented the concept behind it and a rough idea on how we should move about. And, as a result of the discussions, both NIRs and APNIC all agreed that the current fee scheme for NIRs, which is based on per-address fee as well as annual membership fee is not desirable because per-address fee is one-time fee charged for allocations and it's not a stable income for APNIC and it is also not good for budget planning for NIRs. Since this was an informational presentation, we didn't have a clear consensus but this was a thing that we agreed and we also agreed that we will continue discussions on how we will tackle this fee issue in the long run for the NIRs. And we had an update from VNNIC. And we were informed that they have recently revised their fee structure for their own members and also held a second Open Policy Meeting within Vietnam, which is probably like the equivalent of this meeting but in a country-based scale. And unfortunately for the other NIRs, the updates were planned as well but we ran out of time so I would like to apologise to those people who prepared their presentations but could not make this this time and I'll promise that I'll make sure to secure some time for the next meeting. The action items - there are two and the first one is spending from the last meeting and there was a comment to make some revision on ERX FAQ on the current APNIC website and this was from KRNIC and we're going to confirm what particular revision this required. And the second - I think this is what time doing at the moment here - to propose what was reached consensus in the NIR SIG on the abolishing of per-address fee for IPv6 allocations. Any questions so far on the general update? I will go on to explain the details of the proposals but any questions so far? I guess not, so I'll just move on to explain the details. Well, this proposal on changing the fee structure for NIRs which affects APNIC membership as a whole as well NIRs so that is why I would like to explain a little bit more in detail and how it would affect non-APNIC members which are not NIRs. First, for those of you who are not with the idea of per-address fee - this is the type of fee that is charged only to NIRs and every time that NIR receives an allocation from APNIC, the free called per-address fee is charged based on the size of allocations that NIR receives. So as the size of the allocation grows, the charge of per-address fee will also be larger. And I think it would be better to just explain in figures rather than a verbal explanation. So this is how it works. If an NIR receives a /31 allocation, then... well... it's calculated that we will first look at the chart of the utilisation rate for IPv6 which is started for each - for each site which is stated in the policy document so, for /31, it shows that the utilisation of /- no, the number of /48 is nearly 12,000 so what we do is we multiply this 12,417 times the charge for each /48 which is 0.03 US dollars for very large numbers and the charge that - the amount that will be charged at the end would be nearly $752 for a single /31 allocation that an NIR receives. And, on the other hand, for example, for a larger allocation size, for example, a /21, again, we look at the chart that is on the IPv6 policy document and - I'm sorry - I don't have the chart over here but you just have to look at the figures in the utilisation and thank it is nearly 3 million, so you multiply the same amount Ð US$0.03 and the amount would be nearly US$95 million sorry, US$95,000. So this is how the per-address fee works. Well, I think from my explanation - my explanation is not very good but it's quite a complicated way of calculating per-address fee and I'm showing it in diagram how it works. A regular APNIC member simply pays an annual membership fee to APNIC whereas NIRs - we pay the standard annual membership fee and, in addition, we also pay per-address fee under the formula that I have explained right now. And since the amount of per-address fee that is charged in IPv6 to NIRs is very large for when an NIR receives a large allocation, so APNIC has addressed this problem by providing a discount of 90% on when an NIR requests IPv6 allocation, if it is based on the existing IPv4 infrastructure. And, we must note that, however, the minimum charge would be equal to the charge that would be charged for the minimum allocation size which is /32 for IPv6 at the moment. So, if we apply this, what was implemented based on the EC decision, you can see four patterns of fee - just two patterns for each size, for the same size of allocation. For example, if you receive a /3 2 allocation based on IPv6 infrastructure, the charge would be $372 which is the same as what has been charged before. However, if a member requests an allocation request based on IPv4 infrastructure, then a discount of 90% should be implemented which would be $37. However, this is less than the minimum charge that should be applied, which should be US$214, for a /32 allocation, so it shouldn't be charged less than this. That's why the charge, even for /32 would be the same as the charge for - the charge for /31 allocation would be the same as the charge for a /31 allocation which is $214. I think it's really, really, really complicated as you can see. So the problem is it's very, very complicated to calculate this kind of charge and it might cause misunderstanding when a member requests how much they have to pay or they might think, "Oh, you know, I'm suppose supposed to get a 90% discount. Maybe it's actually the case that 100% will be charged" it will be a cause of confusion. So what NIRs thought was, "Let's just make the fee structure simple." So instead of applying this complicated formula, let's completely abolish per-address fee just for IPv6 allocations. Well, the intention behind it is that we don't really want to affect APNIC budget and, you know, we don't want APNIC to run on a red balance,-so-therefore we will keep paying the per-address fee for IPv4 as we have been. And also this proposal is not intended to completely abolish the fee for good. This is a temporary solution until we implement the long-term revision on NIR fee structure, which we should review from scratch. And I'm sure all of you are concerned about the financial impact it would have on APNIC. So we have made a review on this as well and until year 2003, per-address fee charged to NIRs Was almost 0% of APNIC's revenue. And then last year, in year 2004, it was nearly 1% of APNIC's revenue. However, this figure was based on 100% charge. So now 90% discount is applied by EC so we assume the impact on APNIC's budget will be the cut of less than 1% of the total revenue. So that's the idea behind our proposal and to summarise we would like to completely abolish the fee per-address fee, for IPv6 allocations and the reason is because it's very, very complicated and the financial impact on APNIC we believe is not very big, less than 1% of APNIC's revenue and the other note is that this is a temporary solution, so we don't want to leave it completely abolished for good. We're willing to think of a reasonable fee structure that would support APNIC and also take the cost, expenses for IPv6 operations in the long term and the significant has already started discussing on this although it was an informational presentation this time. That's all from me and any questions? RANDY BUSH: Randy Bush, of course. A couple of questions. Slide 12 please. So even though the 95,000 figure looks good, we're talking about 3 million end site customers. You know, I don't think that's much of a burden. I do not find the algorithm complicated, but I kind of want to know if this saving is going to be passed on to me, the LIR. IZUMI OKUTANI: What would be passed on to you? I'm sorry, I couldn't catch it. RANDY BUSH: Do I get the savings? I'm an LIR. I'm getting address space from an NIR. Is the savings being passed on to me? IZUMI OKUTANI: So you're asking, if you are an LIR under NIR would that be passed on to? RANDY BUSH: The cost reduction. IZUMI OKUTANI: OK. In JPNIC's case, that's what we intend to do but I think that would be left up to each NIR's situation. If the operational expenses of NIRs is very big r big and then maybe they can't afford to further cut their fees. RANDY BUSH: But they just got a saving. IZUMI OKUTANI: I guess. BILLY CHEON: Can I answer? When we receive an allocation from APNIC, APNIC is generous with the 90% discount and we exactly applied that discount rate to our members. So I would say yes. IZUMI OKUTANI: And same in JPNIC's case as well. RANDY BUSH: Right. Of course, these are the places that can afford it the most. EUGENE LI: I'm from CNNIC. Basically CNNIC does not charge the LIRs any IPv6 per-address fee. So you know, the members of CNNIC will benefit from this proposal. RANDY BUSH: So they can have 100% discount on nothing. That's a good deal. Lastly I'm confused by the different fees for IPv6 four infrastructure or IPv6 infrastructure when an NIR's infrastructure doesn't exist. IZUMI OKUTANI: Oh, I'm sorry. I wasn't clear on that. I mean when LIRs under NIR management request - make an IPv6 allocation request. So they're not - in the chart - let me show you a diagram. - RANDY BUSH: I understand. But the fee change is from when you get space from APNIC. Where's the infrastructure? IZUMI OKUTANI: So it's based on - do you have (receives laser pointer) Of course, I see you've pointed out that NIRs don't have an IPv6 infrastructure in the same way LIRs do but we submit requests to APNIC for LIRs under our membership so they have IPv6 infrastructure. So we judge it on whether the IPv6 request is based on IPv4 infrastructure or based on IPv6. RANDY BUSH: So you make the request to APNIC one piece for one LIR at a time? IZUMI OKUTANI: Yes. That's right. RANDY BUSH: Got you. Thank you. IZUMI OKUTANI: Any other questions? RAM NARULA: I would suggest if it is possible to come up with a real solution before you abolish the fee scheme. I don't think it's a very good idea to stop something before you come up with something new. Thank you. IZUMI OKUTANI: So your suggestion is that we should abolish it after we come up with, like, a long-term solution. I think the whole point of this is that it takes a long time to implement the long-term solution because the fundamental review of the fee structure would probably expected to take one or two years at least. So until then we have to continue with this really confusing and complicated fee structure. So I understand your comment but that's the situation with NIRs. PAUL WILSON: If there are no more questions then I'd like to put to the meeting the proposal which is that the IPv6 per-address fee for NIRs be completely eliminated as a temporary measure. Is that correct? And ask you to raise your hands if you are in favour. You need to raise them quite high because we will need to count these hands please. OK, and if you are against this proposal, please raise your hand. OK, four against and a considerable number in favour so that's passed by rough consensus and, once again, with an eight-week comment period on the mailing list. OK, thank you. IZUMI OKUTANI: Thank you. APPLAUSE PHILIP SMITH: OK, good afternoon, everybody. The first report I'll cover is from the routing SIG, which had two sessions yesterday. Yesterday? The day before yesterday. From 2:00 to 3:30 and 4:00 to 5:30. Chairs, just as a reminder if you're not aware of, Randy Bush and myself. We had a very full agenda for both sessions. The first session covered two presentations by Geoff Huston, the first on the auto-detection of hijacked prefixes and the work he had been doing there. The second presentation was on AS consumption patterns and finally George Michaelson finished off this session with a presentation and discussion about APNIC resource certification. The second session of the routing SIG followed on after the tea break. It started off with a presentation from Gaurab Raj Uphadaya about analysis work on trying to determine Tier 1 status through ASPATH analysis. Then followed a presentation from Tomoya Yoshida from Japan discussing the deployment of the IRR in Japan and the effect on routing security. I did a short analysis of the status of BGP Flap Damping and where we basically should be going now, basically determining the RIPE document and Randy Bush finished off with allocation and announcement of allocation prefixes, looking at the time interval between prefixes being used and when they first appeared in the routing table. I don't know the exact number in the room. It was probably about half full, so good attendance going by routing SIG past experiences. That was pretty much it. We managed to stick to time, which was quite good. Next was IX SIG yesterday afternoon. We were on the agenda for two sessions but we had a really, really poor response from exchanges and disciplines actually submitting presentations so we only had three items on the agenda when we started. The SIG was chaired by Che-Hoo Cheng and myself. The agenda was pretty much this. (Refers to slide) Nobuhisa Miyake gave one of the regular JPNAP updates and we really appreciate JPNAP's support in the IX SIG. They have been regular presenters. Gaurab gave a regular update on the activities of the Bangladesh Internet Exchange and Nepal Internet Exchange, two of the recent IXes in this part of the world. Tran Kien updated us on the progress of the Vietnam Internet Exchange in Hanoi and Ho Chi Minh City. Then the floor was open for discussions and we had three-quarters of an hour left. Jordi Palet kindly volunteered to spend a few minutes talking about Euro6IX IPv6 exchange in Europe and actually a precursor for the IX SIG meeting at the next APNIC meeting. Jordi has agreed to give a fuller presentation there so we have the first agenda item for the next APNIC meeting. Then we had any other business, a brief discussion about how to go about doing IPv6 address assignments with the Internet exchanges. This was in particular relating to the Nepal exchange. We had - we didn't have the accommodation problem we had at the last meeting and we also didn't have the full agenda problem either. Again, there were people here and listening on the net and if you're interested in Internet exchange connectivities, please volunteer to give a presentation for the next APNIC meeting. That was the IX SIG. Last but not least, the APOPS BoF. APOPS is the Asia Pacific OperatorsÕ Forum. It has existed as a mailing list for many, many years. In fact it started off as a mailing list. A few years ago we started trying to have some kind of formal meeting. Some of you may remember the IX SIG came out of the APOPS BoF event where we had an APOPS meeting discussing Internet exchange points around the region. APOPS has had some use within this community so we still carry on the meeting in some kind of BoF format. Again, with the timing of the meeting, we were squeezed between the regular daily event and the APNIC social event so trying to convince people to come along was hard but we had about 15 or 10. I suppose it is the same issue as everywhere else, trying to arm-twist people into giving presentations. We had three items on the agenda and filled the full hour timeslot we had, which was really good to see. Gaurab, who is the Chair of SANOG, gave a presentation on the last SANOG meeting in Thimpu, Bhutan. Akinori Maemura talked about JANOG activities, particularly the last JANOG in Fukuoka, and we finished off with a nice presentation from Ram Narula, Ôtrivial Internet weaknesses with proposed solutionsÕ. The last presentation brought a lot of really good discussion. I felt unfortunately I had to chop it at the end because we were in danger of missing the bus to go to the APNIC social. I imagine this topic is going to be revisited at APOPS the next time APOPS meets and we look forward to that. So with all these reports I would like to thank everybody who volunteered to present. Thank you to APNIC for supporting the two SIGs and APOPS. Thank you. APPLAUSE PAUL WILSON: Thank you. George, do you have a question? GEORGE MICHAELSON: No, I'm next up. PAUL WILSON: I think you may be. GEORGE MICHAELSON: I have a brief report for the PGP BoF. I think this is the third time we have run a PGP key sharing BoF at an APNIC meeting. This one was very small. We only had four participants. There were so few of us we didn't actually use the meeting room. We just sat around a table and did a simple key exchange. We probably need to do more promotion for APRICOT and it's quite likely with the increased attendance and more of the operations community being there as well as people with an interest in resource management we'll get a larger attendance. It was effective and good to see people getting the keys out there again. That's it. PAUL WILSON: That was short. Phillip has a question. PHILIP SMITH: Not a question, just a comment. Please don't schedule APOPS against PGP because it affects both communities. GEORGE MICHAELSON: That's a good point. The people doing the key signing are inevitably your people. If it doesn't have to be a full BoF, we could happily do it in APOPS if you like. GAURAB RAJ UPADHAYA: The key signing BoF was useful. A lot of people were asking how to use key signing and the BoF was useful. GEORGE MICHAELSON: I think the more promotional message is we should schedule time to give people more information about why they want to be doing this kind of thing. PAUL WILSON: I'd say the APNIC Secretariat, in making the announcement for calls for papers, would be perfectly open to that sort of suggestion even if it is not strictly in the form of a call for papers, Gaurab, so by all means make that suggestion next time. I think there's a few action items with the Secretariat in terms of scheduling some of the SIGs, particularly those who end up with much shorter or longer agendas than originally anticipated. We need to be able to extend into two sessions when needed or collapsed into one, etc. Now I've got on the list here one more BoF, a spam BoF. Is that something that - yes, Suresh has got something to report there. And also the last item is for any other reports so I'm not sure - we didn't have the PGP BoF on here and I'm not sure if I missed out on any others or if anyone had an informal meeting or a nice dinner they would like to report on. Let us know. Suresh. SURESH RAMASUBRAMANIAN: The spam was where I had a few APNIC staffers and a few other people, including Bob Shaw, in the room. We were talking about what kind of demand people have in the spam-related content in these meetings. It was generally agreed that it would - even given the fact it was around 6:00 p.m. and people were mostly making drink and dinner plans by that time, it might have been a good idea to just restrict this to tutorials the next time around instead of having a BoF. We also look for certain things that people operating in the community and region were expecting from spam and suggested ways in which my organization, APCAUCE, could work together with APNIC staff in, shall we say, emphasising the anti-spam-related aspects of some network issues, like why having good DNS and sub-allocating blocks and having good security issues is healthy against spam. That was about it. We also discussed some general cooperation issues. Thanks. PAUL WILSON: Thanks, Suresh. Now I know there was one other BoF which was an ICONS BoF. That was something initiated by the APNIC Secretariat staff who wanted to solicit feedback about ICONS, the ISP support website. I think it was a pretty small affair which is understandable. It is a new project not many people are aware of but we'll try to add to the next APNIC meeting certainly and I encourage anyone interested in that project to have a look. It is linked from the APNIC home page by now, I'm sure. OK, we're just a little ahead of the afternoon tea time. Maybe I could open the mic now over the next five or so minutes in case anyone has a question, suggestion, comment of any kind about the meeting. This is the open mic. OK, to review this afternoon's - the rest of the afternoon then, we've been through the first part of the afternoonÕs business. Afternoon tea is served in just a few minutes. The rest of the afternoon will be a MyAPNIC demonstration showing you the latest with that service and another open mic session. We've got at least 15 minutes allocated there so please maybe have a think over a cup of coffee whether there's something you'd like to raise for the meeting. We've got then some important results. Firstly the Number Council election and then a few lucky draws so there's Nominum lucky draw, there's lucky draw for the winning respondent to the meeting evaluation process, so we'll draw an evaluation form and the lucky person will get something and we also have a winner of the web hunt so there's a lot to look forward to in the rest of the afternoon, isn't there? We'll also be announcing - although it's not really probably not a surprise - where the next meeting will be. I have got a few statistics to show you about who we are here, who has come to the meeting, who we all are, where we come from and so on and that will be about the end of the business. So unless someone wants to approach a microphone right now, let's break for half an hour until 4:00 p.m. and be back here then. OK, thank you. I am informed we have a very small field of contenders for the prize for the evaluation form. The next person to put an entry in will have a 25% chance of winning. I think it is a sensational prize. The next person will have a 25% chance. Anyone afterwards will have a diminishing chance but still a chance. Please give us help with the meeting evaluation. It is part of planning the meetings in the future, how people have found the meeting, whether they have suggestions, constructive criticisms et cetera which can help us construct the meetings in the future. Thank you.