Re: [DNA] Prefix information for link identification in DNA

[Sathya Narayanan <sathya@research.panasonic.com>]

Erik Nordmark wrote:

> *If* we want DNAv6 to handle the case when we don't have global IPv6
> prefixes, I'd suggest we do that by allowing something other than
> prefixes be used as the link identifier (as Keith Moore suggested).

I agree.

*If* we want to allow *something* other than prefix to be used as the
link identifier, the *something* must:

1) be agreed upon by all nodes that will send RA messages.
2) be unique for each link (Weak condition: atleast for links between
which hosts could possibly move).

Since we don't know what this *something* is going to be, we cannot
specify how these two conditions are met, we need to leave that to
future specifications.

AFAICT, the only think we can achieve today is forward compatability for
hosts by defining the syntax for using *something* as link identifier.
As Hesham suggested during the last meeting, syntax can be written as a
TLV and since variable length is allowed by L, new link identifiers will
be possible. In the router advertisement processing text, we should add
a condition to check for link identifier in the RA.

The only question is *whether* we want to allow for this possibility? We
need consensus on this question to proceed further.

Sathya