On Sun, 12 Sep 2010 12:57:48 +0200 (CEST) sth...@nethelp.no wrote: > > >> Serious disconnect between map and reality here, Mark. > > > > > > If you think that, then I don't think you've read my emails > > > properly. > > > > ... or there is a serious disconnect between your reality and my reality. > > > > By saying that L2 devices doing L3 inspection is a layer violation and > > isn't acceptable, you've dismissed the way millions of people are > > connected to the internet as... er, I don't know. Not acceptable and > > wrong. > > Maybe we're disagreeing on the meaning of the word "acceptable". > > I think most of us can agree that L2 devices doing L3 inspection is a > layer violation. > > What a lot of us can *not* agree to is simply dismissing such a feature > *just because it is a layer violation*. > > IPv6 in general is purely an added cost for the providers so far - no > fresh revenue expected in the near future. That means we cannot do large > scale changes of our current architecture. As a consequenc of this, we > need largely the same functionality available for IPv6 as we have for > IPv4 - *even if that fucntionality happens to be a layer violation*. > > We need IPv6 to be "business as usual" as much as possible. A feature > like L2 devices doing L3 inspection is, as far as I can see, a mostly > software thing: Hardware filter to intercept the traffic and route it > to the CPU, probably hardware rate limiting, software doing the actual > inspection. Thus is should be possible to a similar thing for IPv6. >
Unfortunately I know of an instance where the vendor of DSLAM equipment is saying the hardware doesn't have the resources to do these layer 3 inspections for IPv6. So 10s of 1000s of DSLAMs would need to be replaced. Unfortunately, this means the market is going to stay with using PPPoE/PPP despite the overheads. -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
