Re: [Cbor] Supporting IPv6 Link-Local with scope (was Re: Éric Vyncke's Discuss on draft-ietf-cbor-network-addresses-09: (with DISCUSS and COMMENT))

[Brian E Carpenter <[e-mail]>]

On 07-Oct-21 19:43, Jürgen Schönwälder wrote:
> On Thu, Oct 07, 2021 at 03:40:24AM +0200, Carsten Bormann wrote:
>> On 7. Oct 2021, at 00:44, Brian E Carpenter <[e-mail]> wrote:
>>>
>>> That's probably because there are none. As far as I know, there is no
>>> "Scoped architecture for IPv4" RFC. I don't know why there is YANG support.
>>
>> Zones are this gift that keeps on giving.
>>
>> If we are doing zones, we need to do them like YANG.
>> I would have a strong objection against including them but then doing them differently.
>> Maybe we should simply reference RFC 6991 as well.
>>
>> Jürgen:  Do you remember why ipv4-address has a zone ID (called “zone index”)?
>>
> 
> The YANG definitions were derived from the SNMP definitions. I think
> the first RFC having support for zoned IPv4 addresses in the network
> management world was RFC 2851 (published June 2000). I vaguely recall
> that we had several side meetings during the draft writing stage where
> people pointed to other work (that may have led to things like RFC
> 3927 later on) and/or they pointed out that the need to disambiguate
> IP addresses also exists in the IPv4 world. I think people also
> pointed to routers running multiple links using the same non-global
> (e.g., private) address spaces.

Unfortunately that never seems to have led to an equivalent of
RFC4007 or, as far as I can tell, a systematic definition of
interface id or interface index for IPv4, or to any recognition
of all this in the POSIX API. There is just a rather sad statement
in RFC3927: "At the time of this writing, there is no silver bullet
which solves these problems in all cases, in a general way."

The fact that it's covered, in a certain sense, by MIB modules and
YANG modules, doesn't help with this definitional problem. For
the present draft, that is OK, but I think there is a deeper
problem. Probably, we just have to accept that it isn't worth
doing anything about it for the legacy version of IP.

I'm happy with the proposed changes to the draft.

    Brian