Re: [v6ops] Scope of Unique Local IPv6 Unicast Addresses (Fwd: New Version Notification for draft-gont-6man-ipv6-ula-scope-00.txt)

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

On 13-Feb-21 16:17, Fernando Gont wrote:
> On 13/2/21 00:08, Brian E Carpenter wrote:
>> OK, I have deleted all the content (but basically I agree with Ted
>> Lemon). IMHO:
>>
>> ULAs SHOULD be treated exactly like GUAs for all practical purposes
>> (including using a default router for them), with the exception that
>> they MUST be filtered by border routers at a domain boundary that is
>> defined administratively. The only extra requirement is that ULA
>> prefixes MUST be unique within that domain boundary. That's all, I
>> think.
> 
> FWIW, in the context of draft-ietf-6man-slaac-renum, it has been argued 
> that: "only phase out GUAs if you hear of new GUAs, and only phase out 
> ULAs if you hear of new ULAs" -- i.e., that they should be treated 
> differently.
> 
> (Note: I'm not necessarily arguing in favor of this.)

I think it makes sense, since the basic scenario is to have one ULA
for local communication and one GUA for Internet-wide communication.
Having more than one ULA prefix on a link is rather unnatural.
 
>> As for the word "scope" and the phrase "global scope" it may be best
>> to not use them at all. We originally had a false notion of
>> concentric rings of address scope, but reality is a very complex Venn
>> diagram. Even link-local may stop being very meaningful as we see
>> more and more mesh networks. That will be much harder to handle than
>> any aspect of ULAs.
> 
> Deprecate/rev RFC4007? :-)

I think we'd have a lot of documents to touch if we wanted to resolve
this rather than just live with it

   Brian