This archive is retained to ensure existing URLs remain functional. It will not contain any emails sent to this mailing list after July 1, 2024. For all messages, including those sent before and after this date, please visit the new location of the archive at https://mailman.ripe.net/archives/list/address-policy-wg@ripe.net/
[address-policy-wg] IPv6 PI assignment policy
- Previous message (by thread): [address-policy-wg] IPv6 PI assignment policy
- Next message (by thread): [address-policy-wg] IPv6 PI assignment policy
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Maximilian Wilhelm
max at rfc2324.org
Sat Jun 27 16:51:01 CEST 2015
Anno domini 2015 Gert Doering scripsit: Hi, > On Sat, Jun 27, 2015 at 03:29:54PM +0200, Thomas Drewermann wrote: > > the Freifunk communities are not going to give /64 to end users. > > There will be one single IPv6 address leased to end users connecting to > > the wireless networks. > So what's the user to do with this single address, and his network behind > his router? User IPv6 NAT/Masquerading? Hell no :) There is no "his router". > I strongly encourage you to re-think this approach. There seems to be a fundamental misunterstanding: A user in this case is a human using his/her/its mobile/tablet/laptop/youNameIt device and connects it to the wifi network (or connects it via ethernet cable to a network port of a local Freifunk node). It is no intended scenario that anyone connects a router to the Freifunk network to connect own network, so by definition there is no need for NAT. (There are some considerations to used routed /64 inside the Freifunk mesh network instead of a large L2 segment but even then there would be no intended scenario where anyone would connect some non-Freifunk router.) We provide an access network for single devices not for (home) networks and by no means plan on changing that. > [..] > > Since no Freifunk communities has the need for a /32 prefix that would > > be a waste of addresses. > The whole point of IPv6 is to have plenty of addresses - and as there are > 4 billion /32s, using one to give your users at least a /64 is the *right* > way to waste addresses. Do not encourage anyone to use NAT66. This is not an intended scenario for a Freifunk network. We don't want to be in competition with any regular ISP. And we certianly won't encourage anyone to use NAT66. > > @Sascha Luck: I think the policy should reflect that as it does for IPv4. > > Speaking in IPv4 this problem would not have occoured: > > "IP addresses used solely for the connection of an End User to a service > > provider (e.g. point-to-point links) are considered part of the service > > provider's infrastructure." > > > > That problem has already been identified. (page 8) > > https://ripe69.ripe.net/presentations/72-APWG_RS_Feedback_Final.pdf > Yes, we're aware of that, but this is the old "a user only needs > to have a single IP address, and can use NAT" world. Well how about single IP address without NAT? > Since we do not want to encourage this model for IPv6, nobody has ever > brought forward a proposal to allow this approach for IPv6 PI. > (Now, I have no good answer what the Freifunk community *should* do. I can > understand that you're indeed set up quite differently than a traditional > ISP - OTOH, you're not the only one who runs a network on a non-commercial > basis and needs IPv6 addresses. So using PA space from a friendly ISP in > the neighbourhood - like, a /40 or even a /32 - might be a workable > solution... yes, renumbering will be nearly impossible, but right now > the RIPE model doesn't really permit free rides "I want my own addreses, > I want to run something that is similar to an ISP business, I want a slot > in the global routing system, but I am not going to pay for it". We might > want to change our member structure to accomodate non-commercial LIRs - but > that's a topic for the AGM to decide...) IMHO that would be something worth a discussion. Kind regards Max Freifunk Paderborn / Freifunk Rheinland -- If it doesn't work, force it. If it breaks, it needed replacing anyway.
- Previous message (by thread): [address-policy-wg] IPv6 PI assignment policy
- Next message (by thread): [address-policy-wg] IPv6 PI assignment policy
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]