Re: [Roll] Dissenting technical arguments unwelcome

[Richard Kelsey <[e-mail]>]

From: Emmanuel Baccelli [[e-mail]]
Sent: Friday, July 26, 2013 4:50 AM

> As Ulrich mentioned, the conclusion of AUTOCONF was that IP links as
> we know them do *not* make sense in a spontaneous wireless multi-hop
> environment.
> 
> This conclusion was documented in RFC 5889, which essentially bans
> the use of subnet prefixes in this environment. And without subnet
> prefixes, what is the purpose of an IP link? Not much...

Hi Emmanuel,

While I generally agree with what RFC 5889 says, its focus is on
autoconfiguration.  For example, the statement

   o  There is no mechanism to ensure that IPv6 link-local addresses
      are unique across multiple links, hence they can not be used to
      reliably identify routers.

is true only if restricted to autoconfiguration mechanisms.

To the extent that RFC 5889, which is informational, bans anything,
it bans using autoconfiguration to configure addresses with subnet
prefixes in this sort of network.  It says nothing one way or the
other about using, say, DHCPv6 or 6LoWPAN ND to do so.

The whole multi-link subnet discussion is difficult because the term
'subnet' has different implications in different contexts.  For
example, RFC 5889 says:

   Subnet prefix configuration on such interfaces must thus not make any
   promises in terms of direct (one hop) IP connectivity to IP addresses
   other than that of the interface itself.  This suggests the following
   principle:

   o  no two such interfaces in the network should be configured with
      the same subnet prefix.

An equally valid principle would have been:

   o  subnet prefixes should not be used for on-link determination
      for such interfaces.

Which you prefer depends on how much you consider on-link determination
to be a fundamental aspect of subnet prefixes.  For autoconfiguration
it is a fundamental aspect; for routing, perhaps not.

I think it makes sense to follow RFC 5889's recommendation and assign
only /128 subnet prefixes to interfaces in an ad-hoc radio network.  I
also think it makes sense to arrange for those subnet prefixes to share
a single /64 prefix and to use that fact at the routing layer.  Whether
or not this is a multi-link subnet depends on who you ask.

-Richard Kelsey

This message (including any attachments) is intended only for the use of the individual or entity to which it is addressed and may contain information that is non-public, proprietary, privileged, confidential, and exempt from disclosure under applicable law or may constitute as attorney work product.  If you are not the intended recipient, you are hereby notified that any use, dissemination, distribution, or copying of this communication is strictly prohibited. If you have received this communication in error, notify us immediately by telephone and  destroy this message if a facsimile or (ii) delete this message immediately if this is an electronic communication.  
Thank you.