[ppml] Get you IPv6 Today, lets update the policy

[Shane Kerr]

On 2003-01-07 13:46:28 -0800, Klement, Charles wrote:
> 
> Are there currently any plans underway to accommodate the needs of
> Multihomed enterprises?  If my understanding is correct (please
> correct me) The hierarchical nature of IPv6 would not allow me to have
> multiple transit providers.  This makes IPv6 a nonstarter for me.

There are currently two forums where IPv6 multihoming solutions are
being discussed.

One is the IETF Site Multihoming in IPv6 working group:

http://www.ietf.org/html.charters/multi6-charter.html

The IETF working group has been around for a few years, but frankly
hasn't gone forward much.  Which is the reason that a few couple of
interested parties created is the IPv6 Multi-homers list:

http://arneill-py.sacramento.ca.us/ipv6mh/

Based on my understanding of the problem, our current notion of
multihoming is too limited - there will probably be a
"one-size-fits-all" solution to redundant connectivity in IPv6.
Solutions that have good properties for small sites don't support
necessary features for large sites.  This isn't bad - it just means we
may have to stop thinking about "how do I multihome in IPv6" and say
"how do I provide reliablity in IPv6".

Of course, there aren't good answers to that question yet.  :(

As far as I know, you have exactly the same multihoming solution
available now as you do with IPv4 - advertise a more specific route.
Since AFAIK IPv6 folks don't filter heavily on prefix length, this
should give you multihoming today as good as with IPv4.

Yes, it's a suckful situation, but no worse than current practice.

> Is there some loophole where I can qualify as a LIR and "sell" service
> to my remote sites?

I'm not sure what exactly is meant by this.... ???

--
Shane
Speaking only on behalf of myself