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

[Fernando Gont <[e-mail]>]

Hi, Ted,

On 12/2/21 20:29, Ted Lemon wrote:
> On Feb 12, 2021, at 6:12 PM, Fernando Gont <[e-mail] 
> <mailto:[e-mail]>> wrote:
>> Both the uniqueness scope and the routing scope of ULAs are smaller than
>> that of GUAs.
> 
> This is true, but the distinction is that there is no automatic 
> mechanism specified for knowing whether a particular ULA is in scope for 
> a link.

Well, that's because the scope of ULAs is loosely-defined. As Fred, it's 
essentially "site local", with site-local meaning whatever you want it 
to mean (or, IOW, whatever it means based on hor far such packets can be 
routed and the area where such addresses are meaningful).



> It’s perfectly valid to use the default route to forward to a 
> ULA destination. If a host were to automatically not do this, things 
> would break all over the place.

Well, but that's essentially true for anything that is not link-local -- 
because link local is the only scope that is tightly defined.

Global: Valid in all networks
Link-local: only valid within this network segement.

For the rest (e.g. ULAs), you forward them "just in case", because you 
don't really know where the scope ends.




> IOW, what Fred said is perfectly correct and sensible, but the sense in 
> which ULAs are global is that you should (MUST?) use the default route 
> to route them unless there is a more specific route that matches.

This, in away, is the question that this document somewhat raises:

* What's "global scope" anyway?

* And, given what we have right now, either the definition of scope is 
flawed (as in the scoped addressing architecture), or ULAs are not 
global, or both. :-)

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492