comments on Y.Pv6RefModel
I lurk on this list. However I have to say something about this document, mostly because of what’s happening at ITU SG20. It’s taken a lot of work by Marco and others to persuade SG20 to engage with RIPE. And it’s important that we respond clearly. I just have three things to say. 1) The document is very poor. [I’m being uncharacteristically diplomatic.] It contains lots of errors. The proposed addressing plans are just wrong. It would be a *huge* mistake for anyone to adopt these and much, much worse if SG20 recommends them for global adoption. SG20 should abandon this fundamentally flawed document. Work on it simply has to stop. 2) IP addressing plans for IoT devices can’t and shouldn’t be discussed in isolation. They have to be developed in the context/circumstances of the networks where these devices will be used: ie alongside everything else that’s using or will be using the Internet. If members of SG20 have ideas on IP addressing plans, they are more than welcome to come to the RIR fora where their suggestions can be discussed/analysed and possibly incorporated into RIR address policy. That work can’t and shouldn’t take place at SG20. Or elsewhere in the ITU. 3) It’s unacceptable for the ITU to even attempt to get involved in IP addressing. It’s out of scope. They should stick to co-ordination of E.164 numbers and X.25 addresses. IP addressing is primarily a matter for the RIRs. And other Internet-related fora like RIPE, NANOG, IETF, etc. There’s no role for the ITU in this at all.
On 5/25/18 11:07 AM, Jim Reid wrote:
3) It’s unacceptable for the ITU to even attempt to get involved in IP addressing. It’s out of scope. They should stick to co-ordination of E.164 numbers and X.25 addresses. IP addressing is primarily a matter for the RIRs. And other Internet-related fora like RIPE, NANOG, IETF, etc. There’s no role for the ITU in this at all.
hi, about ITU's role, maybe it's worth reading this paper: RESOLUTION 101 (REV. BUSAN, 2014), Internet Protocol-based networks https://www.itu.int/en/action/internet/Documents/Resolution_101_pp14.pdf in that doc it's recalled that: 'the WSIS+10 High-Level Event (Geneva, 2014), in its Statement on the Implementation of WSIS Outcomes and the WSIS Vision Beyond 2015, determined that one of the priority areas to be addressed by the Post-2015 Development Agenda must be: "Encouraging the full deployment of IPv6 to ensure the long-term sustainability of the addressing space, including in light of future developments in the Internet of Things";' and resolves to: 'explore ways and means for greater collaboration and coordination between ITU and relevant organizations [Including, but not limited to, the Internet Corporation for Assigned Names and Numbers (ICANN), the regional Internet registries (RIRs), the Internet Engineering Task Force (IETF), the Internet Society (ISOC) and the World Wide Web Consortium (W3C), on the basis of reciprocity.] involved in the development of IP-based networks and the future Internet, through cooperation agreements, as appropriate, in order to increase the role of ITU in Internet governance so as to ensure maximum benefits to the global community.' so, ITU is just following what the plenipotentiary conference (that is governments, mainly) decided in south korea four years ago. i'd prefer to continue stressing the technical aspects of Y.Pv6RefModel, instead of complaining about ITU's relevance in the topic. thank you -- antonio
Hi Antonio, You're totally right, but precisely because the text that you indicated, the right place for a possible standardization of addressing plans related to IoT, is the IETF, not even the RIRs. So, instead of this document, what the authors need to do is to come to IETF and write an Internet Draft in the appropriate WG. If that WG doesn't exist (just in case), they can call for a BoF to create one. Regards, Jordi -----Mensaje original----- De: ipv6-wg <ipv6-wg-bounces@ripe.net> en nombre de Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> Responder a: Antonio Prado <antonio@prado.it> Fecha: viernes, 25 de mayo de 2018, 13:49 Para: <ipv6-wg@ripe.net> Asunto: Re: [ipv6-wg] comments on Y.Pv6RefModel On 5/25/18 11:07 AM, Jim Reid wrote: > 3) It’s unacceptable for the ITU to even attempt to get involved in IP addressing. It’s out of scope. They should stick to co-ordination of E.164 numbers and X.25 addresses. IP addressing is primarily a matter for the RIRs. And other Internet-related fora like RIPE, NANOG, IETF, etc. There’s no role for the ITU in this at all. hi, about ITU's role, maybe it's worth reading this paper: RESOLUTION 101 (REV. BUSAN, 2014), Internet Protocol-based networks https://www.itu.int/en/action/internet/Documents/Resolution_101_pp14.pdf in that doc it's recalled that: 'the WSIS+10 High-Level Event (Geneva, 2014), in its Statement on the Implementation of WSIS Outcomes and the WSIS Vision Beyond 2015, determined that one of the priority areas to be addressed by the Post-2015 Development Agenda must be: "Encouraging the full deployment of IPv6 to ensure the long-term sustainability of the addressing space, including in light of future developments in the Internet of Things";' and resolves to: 'explore ways and means for greater collaboration and coordination between ITU and relevant organizations [Including, but not limited to, the Internet Corporation for Assigned Names and Numbers (ICANN), the regional Internet registries (RIRs), the Internet Engineering Task Force (IETF), the Internet Society (ISOC) and the World Wide Web Consortium (W3C), on the basis of reciprocity.] involved in the development of IP-based networks and the future Internet, through cooperation agreements, as appropriate, in order to increase the role of ITU in Internet governance so as to ensure maximum benefits to the global community.' so, ITU is just following what the plenipotentiary conference (that is governments, mainly) decided in south korea four years ago. i'd prefer to continue stressing the technical aspects of Y.Pv6RefModel, instead of complaining about ITU's relevance in the topic. thank you -- antonio ********************************************** IPv4 is over Are you ready for the new Internet ? http://www.consulintel.es The IPv6 Company This electronic message contains information which may be privileged or confidential. The information is intended to be for the exclusive use of the individual(s) named above and further non-explicilty authorized disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited and will be considered a criminal offense. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited, will be considered a criminal offense, so you must reply to the original sender to inform about this communication and delete it.
On 25 May 2018, at 12:49, Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> wrote:
i'd prefer to continue stressing the technical aspects of Y.Pv6RefModel, instead of complaining about ITU's relevance in the topic.
Antonio, while it’s important to talk about the technical aspects of Y.IPv6RefModel -- please tell the WG what you think about it! -- we should not lose sight of the layer 9+ aspects. These are important too. SG20 didn’t volunteer to bring this document to the WG. They were made to do it. If those interventions in Geneva hadn’t happened, we probably wouldn’t have seen this document until it finally got published as a Recommendation.
On 5/25/18 3:54 PM, Jim Reid wrote:
we should not lose sight of the layer 9+ aspects. These are important too.
i agree but i believe there are more suitable locations for issues concerning those aspects. thank you -- antonio
On 5/25/18 3:54 PM, Jim Reid wrote:
Antonio, while it’s important to talk about the technical aspects of Y.IPv6RefModel -- please tell the WG what you think about it! hi,
i with pleasure accept your kind invite. it's clear to everybody now that RIPE's community opinions on Y.IPv6RefModel are not positive, to be fair. anyway, about the procedural side: ITU commissioned a research to Mandat International, UN consultant, to follow what their resolution 101 decided in 2014. the SG20 in 2018 recommended on the draft that: "The cost/benefit analyses of these approaches and addressing schemes will need careful consideration. That work should be developed in consultation with those who have operational expertise in running such networks vis [via] the RIR policy making fora." RIPE NCC, as a sector member, will send our comments to SG20 (Marco Hogewoning in marseille said). somehow here we are. about the model proposed in the draft, i read two quotes from RFC 6540 and an IAB statement (page 11): "all networking standards assume the use of IPv6, and be written so they do not require IPv4". i think we all agree with it, but at this point i cannot understand why the draft insists on IPv4 by saying (page 15): "To enable mapping between and consistency between IPv4 and IPv6 subnet addressing plans, a dual strategy is proposed, with part of the IPv6 subnet addressing plan designed to map corresponding IPv4 addresses, with the possibility to extend the IPv6 subnet addressing plan and benefit from its scalability where this constraint is not required." to me, this is an enormous contradiction of principle that makes the whole work vanish at once. finally, i agree with all other tech arguments brought up in this mailing list earlier. thank you -- antonio
I already said this before, but "... in running such networks vis [via] the RIR policy making fora." is wrong. When discussing about addressing models, the right expertise is in IETF. I'm not saying the RIRs don't have that expertise, and in fact many folks with the right expertise is working in both sides (IETF and RIRs). The point is that this looks like a model for addressing, with the aim of standardizing it, and the most correct WGs is probably v6ops (operation of IPv6 networks), or even 6man, which the cooperation of IoT WGs. An example of this is obviously RFC4291. The RIRs policies don't work in addressing models or plans, in fact, the actual IPv6 policy say: "The middle bits of an address indicate the subnet ID. This field may often be inefficiently utilised, but the operational benefits of a consistent width subnet field were deemed to be outweigh the drawbacks. This is a variable length field, determined by each LIR's local assignment policy." In other RIRs this is even more clear, they often have a sentence like: "The exact size of the assignment is an operational decision for the LIR or ISP to make." Of course, I'm not saying that we can't change our policies, but I've the feeling that the RIRs communities will not reach consensus in something as this document, unless the IETF also support it. Regards, Jordi -----Mensaje original----- De: ipv6-wg <ipv6-wg-bounces@ripe.net> en nombre de Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> Responder a: Antonio Prado <antonio@prado.it> Fecha: sábado, 26 de mayo de 2018, 12:28 Para: Jim Reid <jim@rfc1035.com> CC: <ipv6-wg@ripe.net> Asunto: Re: [ipv6-wg] comments on Y.Pv6RefModel On 5/25/18 3:54 PM, Jim Reid wrote: > Antonio, while it’s important to talk about the technical aspects of Y.IPv6RefModel -- please tell the WG what you think about it! hi, i with pleasure accept your kind invite. it's clear to everybody now that RIPE's community opinions on Y.IPv6RefModel are not positive, to be fair. anyway, about the procedural side: ITU commissioned a research to Mandat International, UN consultant, to follow what their resolution 101 decided in 2014. the SG20 in 2018 recommended on the draft that: "The cost/benefit analyses of these approaches and addressing schemes will need careful consideration. That work should be developed in consultation with those who have operational expertise in running such networks vis [via] the RIR policy making fora." RIPE NCC, as a sector member, will send our comments to SG20 (Marco Hogewoning in marseille said). somehow here we are. about the model proposed in the draft, i read two quotes from RFC 6540 and an IAB statement (page 11): "all networking standards assume the use of IPv6, and be written so they do not require IPv4". i think we all agree with it, but at this point i cannot understand why the draft insists on IPv4 by saying (page 15): "To enable mapping between and consistency between IPv4 and IPv6 subnet addressing plans, a dual strategy is proposed, with part of the IPv6 subnet addressing plan designed to map corresponding IPv4 addresses, with the possibility to extend the IPv6 subnet addressing plan and benefit from its scalability where this constraint is not required." to me, this is an enormous contradiction of principle that makes the whole work vanish at once. finally, i agree with all other tech arguments brought up in this mailing list earlier. thank you -- antonio ********************************************** IPv4 is over Are you ready for the new Internet ? http://www.consulintel.es The IPv6 Company This electronic message contains information which may be privileged or confidential. The information is intended to be for the exclusive use of the individual(s) named above and further non-explicilty authorized disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited and will be considered a criminal offense. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited, will be considered a criminal offense, so you must reply to the original sender to inform about this communication and delete it.
On 25 May 2018, at 13:49, Antonio Prado via ipv6-wg wrote:
'explore ways and means for greater collaboration and coordination between ITU and relevant organizations [Including, but not limited to, the Internet Corporation for Assigned Names and Numbers (ICANN), the regional Internet registries (RIRs), the Internet Engineering Task Force (IETF), the Internet Society (ISOC) and the World Wide Web Consortium (W3C), on the basis of reciprocity.] involved in the development of IP-based networks and the future Internet, through cooperation agreements, as appropriate, in order to increase the role of ITU in Internet governance so as to ensure maximum benefits to the global community.'
so, ITU is just following what the plenipotentiary conference (that is governments, mainly) decided in south korea four years ago.
No, the text above, where the names of he organisations are in a footnote, implies precisely that ITU should *not* develop their own things but instead cooperate so that what is developed in other SDOs matches whatever needs ITU believes exists. Without the text, RIPE would never have got the liaison in the first place. See you in Dubai. Patrik
On 5/25/18 4:41 PM, Patrik Fältström wrote:
No, the text above, where the names of he organisations are in a footnote, implies precisely that ITU should *not* develop their own things but instead cooperate so that what is developed in other SDOs matches whatever needs ITU believes exists. Without the text, RIPE would never have got the liaison in the first place.
hi, you clearly say it implies that, ok. i'm just stressing the fact that in this wg it's worthless arguing about ITU's manners as a further argument in favor of the lack of technical quality of the paper. let's stick to the tech issues. thank you -- antonio
In addition to agreeing to all of what you wrote: Jim Reid wrote:
1) The document is very poor. [I’m being uncharacteristically diplomatic.] It contains lots of errors. The proposed addressing plans are just wrong.
"Arbitrary" would be a more appropriate term than "just wrong" in this context. There is no objective basis for carving ipv6 site allocations into the 5 categories specified. It is one way of doing things, but there are many others and there is no fundamental basis for the ITU to recommend this model over any other. Tying ipv4 and ipv6 allocation strategies together is bizarre and in my experience, pointless to the degree of being self-destructive. IPv4 suffers from potentially crippling shortages and address allocation optimisation requirements for ipv4 bear no relation to sensible and relevant optimisation strategies for ipv6. Is is extraordinary to see the two conflated in a document like this.
It would be a*huge* mistake for anyone to adopt these and much, much worse if SG20 recommends them for global adoption. SG20 should abandon this fundamentally flawed document. Work on it simply has to stop.
It's not good enough to shout at the ITU and say it's out of scope - as Antonio noted, this work is going to go ahead at the ITU whether it makes sense or not. Probably the RIR / NRO community needs to examine this problem itself and either come up with a series of recommendations or non-recommendations. Nick
participants (5)
-
Antonio Prado -
Jim Reid -
JORDI PALET MARTINEZ -
Nick Hilliard -
Patrik Fältström