Re: Fwd: Re: note from the iesg plenary

[Randy Bush <randy@psg.com>]

> (3) IPv6 anycast is substantially different from the anycast hacks
> currently used in IPv4.

this is not generally appreciated, in both senses of the word.

> One big difference: a global IPv6 anycast address would (probably?) not
> appear to be inside of anyone's AS.

this part i missed.  or by 'AS' do you mean OSPF area?  in this gang, 'AS'
is a bgp thing.

> And, despite the fact that we are specifying the use of site-local
> anycast addresses in documents (like Dave Thaler's DNS stuff), we
> don't actually know what a site-local anycast would look like.

why would those of us who want an anycast interface to appear to the world
as no different from a unicast interface not think it should be the same as
a site local unicast anycast address?

> However, the IPv4 stuff has only been tried with BGP.

not true at all.

> Do we know for sure that the OSPF tree will converge if it thinks that
> one host is actually located at two different places in the topology?

hmmm.  i will admit to not having checked the multi-area issue.

> I'm glad that someone was listening...  :-).

my apologies for not listening well enough.

> You and Steve resolved this case by stating that a site must be
> "convex", and I am sure that the text explaining this will be in the next 
> version of the document -- I do understand what it means, and I agree
> that it will solve the problem.
> 
> However, this does place several constraints on the concept of a site.
> For example, consider two office of the same company, both connected to 
> the Internet, that are connected by a high-cost link (higher cost than
> traversing the Internet between the two locations).  These two 
> locations could not be configured as a single site, although they 
> would probably cross-pollinate their global addresses (for redundancy)
> and might be treated as a single routing AS.  This adds some complexity
> for routing purposes, as it clearly decouples the concept of sites
> (which must be convex) and routing ASes (which do not have that
> restriction). 

i think things are getting very crufty here, and we will regret trying to
create these complex little sub-definitions of site etc.

> Just for the record, I think that I know the answers to many of the
> questions that I have posed.  I'm sure that several of you know the
> answers as well.  But, do we know the same answers?

and, as you say, have these questions and their answers been clearly and
simply spelled out in the published specs?

From: "Brian Haberman" <haberman@nortelnetworks.com>
> From discussions I have had, I don't think there is much interest
> in global anycast.  The proposals I have heard have been for using
> anycast within domains, so that the anycast address gets aggregated
> into the global prefix for that domain.

aiiiii!!!!!  global v4 anycast is used significantly.  let's not further
break v6 anycast, please.

> In your scenario, if the sites had a cheaper connection via the Internet,
> I would not even consider putting them in one site.

what is 'putting them in one site'?  do you mean considering them as a site
in terms of things such as site-local addressing/routing?  or are you
starting to mandate operational restrictions?

> If it was necessary to have both locations in the same site, it would
> still be possible with the use of IPSec tunnels and some tweaked routing
> parameters.

please explain.

randy