On 2011-02-17 03:47, Livingood, Jason wrote:
Parts of the challenge here is that turning on IPv6 (publishing a AAAA)
can also cause brokenness for users that have no IPv6 connectivity, e.g.,
those relying on broken 6to4 relays. This has been documented all over
the place, for example here:
<http://ripe61.ripe.net/presentations/162-ripe61.pdf>
So even if there are very few IPv6 eyeballs, this event can serve to
flush out that flavor of brokenness. As I understand it, part of the
idea of everyone moving together is to get people to see the brokenness
across multiple sites, thus to blame the network not the content
provider, thus to pressure the networks to fix things.
Richard is exactly right on where a lot of value is. This is an
opportunity to find and fix the ~0.05% level of brokenness. Even
"non-participating" ISPs will need to take steps to prepare, and this is
of course a great forcing function within companies to ask what their IPv6
plans and to begin/continue IPv6 technical training, etc.
Over in v6ops, we have had some vigorous discussion about the anycast 6to4
brokenness and there is a draft:
draft-carpenter-v6ops-6to4-teredo-advisory
My hope is that this will be in good enough shape prior to June 8th
that it can contribute to the day. Discussion welcome on the v6ops list.
Brian
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf