Hi Leo, Yes, good point, that's the idea. Regards, Jordi

De: Leo Vegoda <leo.vegoda@icann.org> Responder a: <leo.vegoda@icann.org> Fecha: Fri, 26 Jan 2007 10:01:30 +0100 Para: <jordi.palet@consulintel.es> CC: "address-policy-wg@ripe.net" <address-policy-wg@ripe.net> Asunto: Re: [address-policy-wg] Concerns on policy proposal 2005-08

Jordi,

On Jan 25, 2007, at 4:41 PM, JORDI PALET MARTINEZ wrote:

[...]

...

5.4.1. I think is a mistake to change the existing text, but if we decide to go for it, I will never suggest a minimum value of /64. End-users need to have at least a prefix that allow to have several subnets. My view is still that /48 is the right size, but definitively I will accept in the *worst* case something such as /56, but never /64 should be suggested as a minimum value. I think if we go into this direction, definitively there is a need to clearly state that the end-users have the right to request extra size if needed, without any additional justification,

When you say that end users should have the right to "request" extra space if needed "without any additional justification", do you actually mean that end users should have the right to *receive* extra space without any additional justification?

Regards,

-- Leo Vegoda IANA Numbers Liaison

********************************************** The IPv6 Portal: http://www.ipv6tf.org Bye 6Bone. Hi, IPv6 ! http://www.ipv6day.org This electronic message contains information which may be privileged or confidential. The information is intended to be for the use of the individual(s) named above. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information, including attached files, is prohibited.

Leo Vegoda wrote:

Hi Jordi,

On Jan 26, 2007, at 10:04 AM, JORDI PALET MARTINEZ wrote:

...

Hi Leo,

Yes, good point, that's the idea.

Are you trying to limit the commercial freedom of ISPs to offer a range of products based around different sized networks?

Well, I think it is inevitabale to revisit that issue and to reopen that case. For 2 reasons: - address space was never to be "sold" to end users by "volume", it was always to be provided on an as needed basis. Address space is not an ISPs "property" to be resold for profit. There are many other (better) ways to implement commercial freedom and product differentiation. In the IPv4 world, the (marketing departments of) ISPs got away with (mis)using the scarcity argument for wiring the size of address blocks into the packages they are offering. My personal view on that: the users got used to it and the community, including the RIRs were/are tolerating this approach. - for IPv6, one of the explicit goals (architectural assumptions) was that the community should do away with the bickering, the mechanisms that support or allow IPv4-like ISP lock-in mechanisms, by *assigning* equally sized blocks (/48) to end sites because IPv6 addresses are plenty. Other than trying to limit the attractiveness of "selling" addresses, there are more reasons why a homogenous address distribution seems favourable: When moving site networks around (ISP a -> ISP b) the site would not have to re-design the subnet layout, just the LHS bits in the addresses would have to be changed. (Anyone still remembering A6 RRs to make that easy? ;-) ) Now if/when we encourage the situation that different ISPs offer different "packages", it is easy to end up in a situation where a site would be holding a say /48, but another ISP would (by default) only assign /56. Then, either the site would have to re-structure its subnet landscape or *and that's the point here* should have the right to request *and* to receive an equally sized and *contiguous* assignment from ISP b, up to the originally proposed /48. Anythig else is (again, and through the same backdoor of "conservation") just a motivation again to "sell" address space again. And to promote address translation mechanisms and products...

I remember this issue being discussed in the context of IPv4 when ripe-152, the 'Charging by Local Internet Registries' document was reviewed a couple of years ago. I seem to remember several strong statements against RIR policy regulating business models during that discussion. Is this a discussion you want to re-visit?

In the context of this proposal, I think we need to do that, otherwise we are pulling the carpet away from underneath the original architectural assumptions and recommendations for IPv6.

http://www.ripe.net/ripe/wg/address-policy/r47-minutes.html (Section F)

Regards,

The problem(?) that I'm seeing here is that the RIPE community is made up of more ISPs than end-users and customers. So there may well be a bias to preserve the IPv4 business models for the IPv6 world. So, when you are asking "Are you trying to limit the commercial freedom of ISPs", from a communities point of view and my personal one: yes, *in this particular respect* because there are plenty of other and more appropriate mechanisms to use for product differentiation. Wilfried

On Jan 26, 2007, at 11:51 AM, Wilfried Woeber, UniVie/ACOnet wrote: [...]

...

Are you trying to limit the commercial freedom of ISPs to offer a range of products based around different sized networks?

Well, I think it is inevitabale to revisit that issue and to reopen that case. For 2 reasons:

- address space was never to be "sold" to end users by "volume", it was always to be provided on an as needed basis. Address space is not an ISPs "property" to be resold for profit. There are many other (better) ways to implement commercial freedom and product differentiation.

In the IPv4 world, the (marketing departments of) ISPs got away with (mis)using the scarcity argument for wiring the size of address blocks into the packages they are offering.

My personal view on that: the users got used to it and the community, including the RIRs were/are tolerating this approach.

There are more points to consider. If you have a policy that requires an LIR to always assign a /48 on any product at the request of the end user you take away the ability of the LIR to plan their utilisation rate for any network segment or their allocation overall. We need to be careful not to set up a situation where an LIR has to achieve a particular assignment density to qualify for a new allocation but cannot do this because customers are always entitled to a /48 rather than a /56. Just as importantly though, you create a policy without an enforceable appeal process or enforcement mechanism. That puts RIPE NCC staff in an untenable position where they have to implement a policy that is designed to see LIRs come back for address space very infrequently (see: 3.7) but does not give them the tools to enforce this right to always receive a /48. Regards, -- Leo Vegoda IANA Numbers Liaison

Hi, On Fri, 26 Jan 2007, JORDI PALET MARTINEZ wrote: (...)

What I recall is that address space is not a good to be sold by the LIRs, and they don't own it. Only the real cost for the "management" should be charged, and I can tell you for sure that this is not in the order of 12 Euros per month per IPv4 address as it is being charged in Spain (I guess about the same in other countries).

Yes. Unfortunately :-( If an end-customer wants a fixed public IP, it usually pays for an "extra".

So the real situation with IPv4 is that this is not the case. Many ISPs actually "sell" the addresses.

Yes. Afaik, this is true for ISPs with a broad spectrum of customers (but not every LIR is doing this...). Theory <> Practice. :-(

We need to understand that if we do the same with IPv6, we in fact kill IPv6

Kill "early deployment of" IPv6. In five/six years from now, global needs should be different... ;-)

and there is no advantage to deploy it. The possible business with IPv6 is the fact that new services and applications can be generated and make profit

Which new services and applications? 99,9999... % of engineers/programmers/designers/... are "formatted" for the v4-only perspective. And i'm not even thinking about the holy grail - the killer app - but the fact is that there's still no application that could *significantly* work better in v6 when compared to v4.

from them BECAUSE: 1) there are enough IPv6 addresses 2) its provision is easier (specially if all the end-users get the same prefix, so network become "flat", such as /48 right now).

Very marginal effect on profit. This can in fact help reduce expenses, but no real effect on profit.

3) there is no NAT (which will be better avoid if we have 1 and 2 above)

Unfortunately most people like NAT. :-(

Probably we can add other things to the list, but the important point is to make sure that the addresses are there for end-users applications and services at no cost (and I'm not talking just about money, but "justification" cost).

Justification cost = Time is money! :-) I can partially agree with this. However, on the other side i can also see that bigger ISPs could get an eye on IPv6, if they could make some extra money by assigning /48s to customers. Smaller ISPs could of course get into IPv6 from the perspective of offering something others still don't. Cheers, ------------------------------------------------------------------------- Carlos Friac,as See: Wide Area Network Working Group (WAN) www.gigapix.pt FCCN - Fundacao para a Computacao Cientifica Nacional www.ipv6.eu Av. do Brasil, n.101 www.6diss.org 1700-066 Lisboa www.geant2.net Tel: +351 218440100 Fax: +351 218472167 www.fccn.pt ------------------------------------------------------------------------- "Internet is just routes (209927/730), naming (billions) and... people!" Aviso de Confidencialidade Esta mensagem e' exclusivamente destinada ao seu destinatario, podendo conter informacao CONFIDENCIAL, cuja divulgacao esta' expressamente vedada nos termos da lei. Caso tenha recepcionado indevidamente esta mensagem, solicitamos-lhe que nos comunique esse mesmo facto por esta via ou para o telefone +351 218440100 devendo apagar o seu conteudo de imediato. Warning This message is intended exclusively for its addressee. It may contain CONFIDENTIAL information protected by law. If this message has been received by error, please notify us via e-mail or by telephone +351 218440100 and delete it immediately.