[atnog] Wiederbelebung der österreichischen IPv6 Taskforce

Mike Knell m at blat.at
Thu Mar 11 17:43:02 CET 2021

> On 11.03.2021, at 17:04, Kaiser, Karl <karl.kaiser at magenta.at> wrote:
> Hallo,
> prinzipiell bin ich bei dir – v6 in Österreich weiter voranzutreiben ist generell die Richtung in die wohl alle in dieser Gruppe wollen.
> Es gibt wohl inzwischen hierzulande keinen ISP, der nicht in der Lage und auch Willens wäre, einem Kunden IPv6 zu liefern.
> Allerdings sehe ich vor allem im B2B Bereich keine große Nachfrage – solange man noch genug v4 Adressen zur Verfügung hat.

Ah, this is a very familiar argument - I followed the IPnG stuff in the 90s and was involved in the very early 6BONE experiments in 2000 or so, connecting my then university employer to the v6 network (via a router PC running one of the only IPv6 stacks then available - FreeBSD). It worked fine then, and although in 2000 the idea of IPv4 exhaustion was “something that will happen in the future which we should prepare for” rather than “something that has already happen but which we’re pretending hasn’t” it’s deeply depressing and hard to believe how little attention the Internet at large has paid to this issue since then. Instead, NAT was developed as a temporary stopgap (!), then mis-sold as a security tool, and now I just keep seeing the same arguments being churned out over and over again about how some form of address translation is essential and desirable “because security”.

The situation as it stands is extra-depressing because it neglects one of the founding principles of the Internet - traffic agnosticism and accessibility. Since networking became a purely commercial transaction this has become gradually worse - for instance, with my home connection to UPC I can get a DS-Lite connection (which breaks a number of things) or I can get a single static v4 address (which breaks other things).

I know that selling proper IPv6 coverage to the suits is not easy because they see CG-NAT (one of the most destructive innovations to hit end-to-end networking transparency ever) and bidding stupid amounts for blocks of dubiously resellable v4 as a more cost-effective transaction than investing the engineering time and hardware needed to do a proper job on v6. But sticking to half-assed solutions will make nobody happy in the long run. While stuff like DS-Lite is presenting what is effectively a distorted version of IPv6 to end users, v6 will remain an area where FUD means “oh, if you have networking problems it’s probably IPv6”. But this is nothing new - in 2000, where institutional web proxy clusters became a thing due to demand outstripping bandwidth, everything that went wrong got blamed on the proxy too. :)

So, I guess I’m only here as a spectator (work’s production connectivity is ably taken care of for us by people cleverer than me) but as a *user* of the Internet the fact that I can’t get proper, dual-stack v4/v6 connectivity at my home router is just grim. Until something is done about this, the democratic hole in the Internet will remain as globally routable address space will continue to be something affordable only to large companies willing to bid three-figure sums per address.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL: <http://atnog.at/pipermail/atnog/attachments/20210311/6da21eb9/attachment.sig>

More information about the atnog mailing list