[AusNOG] Prediction: Google et. al. may artificially penalise IPv4 clients

[Mark Andrews]

In message <ED49A80A-7137-4043-8C12-EC0E2AB51C3D at gmail.com>, Tim Raphael writes
:
> So TLDR: Lots of fundamental things are broken.
>
> There is no point waiting on the standards bodies as it will be an
> eternity until all of this is addressed so what is the pragmatic way
> forward?

This isn't a standards body issue.  This is you guys requesting
WORKING IPv6 equipement and sending it back when IT DOES NOT WORK.

If your router can't route packet with a EH or a fragmentation
header present then it is a PoS and should be return AS NOT FIT FOR
PURPOSE.  Sorry you guys are buying this garbage and as long as you
continue to do so things won't get better.

If you load balancer can't look inside the ICMPv6 header to workout
where to deliver the ICMPv6 PTB message it should be returned to
the supplier as NOT FIT FOR PURPOSE.

It shouldn't require a RFC to say "Look inside the ICMP packet and
work out from the embedded addresses where to deliver the ICMP
message".

> In this scenario, I can well understand smaller networks being cautious
> but there are plenty of small and medium networks using v6 just fine at
> the moment.
> Perhaps they dont feel the same pain as they take the simpler approach
> and not buy into some of the currently / apparently broken or more
> complex technologies.
>
> Either way, it still sounds like we have a very long way to do.
>
> - Tim
>
>
> > On 2 May 2017, at 9:37 am, Geoff Huston <gih at apnic.net> wrote:
> >
> >
> >> On 2 May 2017, at 8:20 am, Tim Raphael <raphael.timothy at gmail.com>
> wrote:
> >>
> >> Given its been a while since our last discussion what are the
> remaining roadblocks for ISPs taking up v6?
> >>
> >> Lack of funds is not a valid point.
> >>
> >> - Tim
> >>
> >
> > You asked.
> >
> > IPv6 Packet Fragmentation handling is broken.
> >
> > Equally, IPv6 Extension Header handling is a mess.
> >
> > SLAAC and RDNSS and DHCP6 is a chaotic mess.
> >
> > Routing coverage is erratic.
> >
> > As long as you treat IPv6 kindly, use very small TCP MSS values, avoid
> UDP, pay care in delivering all ICMPv6 messages, run host MSS caches af
> necessary, and be ultra careful about routing both in terms of the
> completeness of the routes you see and the extent to which your routes
> get propagated then IPv6 will likely work just fine once youve got past
> the auto-config stage. But push it harder in any of these areas and
> you'll regret it.
> >
> > So I suspect for many folk it requires a level of loving care and
> technical attention that they are not willing or able to expend right
> now. I can well understand that, and Im not willing to castigate anyone
> for still being cautious. Whether you deploy IPv6 today or tomorrow you
> cant stop the ongoing task of cramming more stuff into IPv4 just yet. So
> if your engineering and tech assistance resources are limited, I can
> readily understand anyone who is still waiting just a bit longer. With a
> bit of luck we'll find solid ways to avoid these issues in the coming
> months and the entire task will then be a lot less forbidding than it is
> now. Or we wont. Can't tell yet.
> >
> > Geoff
> >
> >
> >
> >
> >
>
> _______________________________________________
> AusNOG mailing list
> AusNOG at lists.ausnog.net
> http://lists.ausnog.net/mailman/listinfo/ausnog

-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org