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

["Templin (US), Fred L" <[e-mail]>]

David, independent of how “scope” is defined, nodes that connect to MANETs that
have no infrastructure connections will need to use some kind of IPv6 address for
MANET-local communications. LLAs are out, because they can only travel 1 hop and
MANETs are by definition multihop. ULA-Ds are possible, but when a hodgepodge
of nodes come together to form a MANET there is no guarantee they will all share
a common Global-ID and Subnet-ID. ULA-Cs are a mystery because no one has
defined how to use that space yet. So, that leaves HITs which by their nature are
not aggregatable beyond /128. In that way, HITs look like GUAs but they are in
some ways of a lesser “scope” than (aggregatable) GUAs – so it seems like there
should be some way to capture that.

Fred

From: David Farmer [mailto:[e-mail]]
Sent: Thursday, February 18, 2021 2:52 PM
To: Templin (US), Fred L <[e-mail]>
Cc: Fernando Gont <[e-mail]>; IPv6 Operations <[e-mail]>; [e-mail]
Subject: [EXTERNAL] Re: [v6ops] Scope of Unique Local IPv6 Unicast Addresses (Fwd: New Version Notification for draft-gont-6man-ipv6-ula-scope-00.txt)


EXT email: be mindful of links/attachments.




So let's be clear "IPv6 Address Scope" does not really have the general meaning of the word "scope", it has a very specific meaning that in my opinion boils down to are Zone Indexes or Identifiers needed to uniquely identify them.

There is an argument to be made that ULAs should be Scoped, and if they were globally reachable they would have to be Scoped. Nevertheless, they are not globally reachable, and therefore so while it is possible that there is another user of the same ULA prefix you are using. But, if you follow the algorithm, in RFC4193 for allocation, the probability of that duplicate user ever being reachable by you is very-very small.

So then do you want to use Zone Indexes or Identifiers with MANET? I doubt it, personally, I believe that would effectively make MANET unusable by most users.

Thanks

On Thu, Feb 18, 2021 at 2:11 PM Templin (US), Fred L <[e-mail]<mailto:[e-mail]>> wrote:
Fernando,

Then, let's have a scope for "MANET-local" scope which is what you get when
you have a small collection of (probably) mobile nodes that form a network
between themselves without the support of any infrastructure. I think HITs
would make for fine IPv6 addresses within the MANET-local scope, but if the
MANET ever connects to the global Internet then nodes will also want to get
a *real* GUA (served by the infrastructure) to go along with the HIT they
already have.

Did I mention HITs are self-generated, and not delegated by some
infrastructure node?

Fred

> -----Original Message-----
> From: Fernando Gont [mailto:[e-mail]<mailto:[e-mail]>]
> Sent: Thursday, February 18, 2021 11:53 AM
> To: Templin (US), Fred L <[e-mail]<mailto:[e-mail]>>; IPv6 Operations <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>
> Subject: Re: [v6ops] Scope of Unique Local IPv6 Unicast Addresses (Fwd: New Version Notification for draft-gont-6man-
> ipv6-ula-scope-00.txt)
>
>
> On 18/2/21 16:46, Templin (US), Fred L wrote:
> [...]
> >>
> >> Hi, Fred,
> >>
> >> On 18/2/21 16:11, Templin (US), Fred L wrote:
> >> [...]
> >>>> On 18/2/21 14:37, Templin (US), Fred L wrote:
> >>>>> Another aspect I failed to mention is that the use of (H)HITs does not
> >>>>> necessarily  mean that all aspects of the HIP protocol must be used. (H)HITs could be
> >>>>> used with the AERO/OMNI services instead, for example.
> >>>>
> >>>> Questions: Are these addresses globally-unique?
> >>>
> >>> Yes - global uniqueness is a key objective of (H)HIT. Aggregation is not within
> >>> scope, however.
> >>
> >> If that's the case, then, according to RFC4007, they are global addresses.
> >
> > I actually think HITs are a bit of a gray area. According to RFC7343, they are
> > formed from the GUA prefix 2001:20::/28 but the remaining bits following
> > the ::/28 prefix are cryptographically generated and hence non-aggregatable.
>
> BUt according to RFC4007, the scope is defined by the topological span
> where the address is meaningful....
>
>
> > That means that there is no way to represent a group of them in the routing
> > system using any other prefix length than /128.
>
> That's not a requirement when it comes to "scope".
>
>
>
> >> (I'd argue that if they cannot be aggregated, that's because they
> >> resulting "addresses" are not really topologically-dependent, in which
> >> case you might probably argue that they are not addresses in the first
> >> place :-) )
> >
> > Not in terms of routing across the global Internet, no, because it would not
> > scale to inject large numbers of /128's into the global Internet routing system.
>
> The thing is that if they don't represent a location in the network
> topology, they don't seem to qualify as an address, anyway.
>
> An address is supposed to be a topologically-dependent identifier --
> i.e., to convey information of location ("where?").
>
> Thanks,
> --
> Fernando Gont
> SI6 Networks
> e-mail: [e-mail]<mailto:[e-mail]>
> PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
>
>
>

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[e-mail]<mailto:[e-mail]>
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------


--
===============================================
David Farmer               Email:[e-mail]<mailto:[e-mail]>
Networking & Telecommunication Services
Office of Information Technology
University of Minnesota
2218 University Ave SE        Phone: [phone]
Minneapolis, MN 55414-3029   Cell: 612-812-9952
===============================================