Re: <draft-ietf-6man-rfc4291bis-09.txt>

[james woodyatt <[e-mail]>]

On Jul 24, 2017, at 10:47, t.petch <[e-mail]> wrote:
> 
> I think I'd prefer to call the righthand side of both a subnet prefix and a on-link prefix an interface identifier, because in both case it is a node’s interface that is being identified.  Then generically they are interface identifiers, if you are referring to a specific aspect it is qualified with the aspect.

Here is the reason I can’t do that: the fact a destination address has an on-link prefix does not, in fact, identify a destination at a node on that link. For one, it may be an anycast address. However, even if it’s a unicast address, it may— in practice, if not in standard theory— identify a destination on a node that in reality is attached to a completely different link.

An important exception, with which I find myself increasingly concerned these days, is the case where a 6LBR is sharing the subnet prefix advertised in PIO by a CE router, preferably, with plen=64, L=1 and A=1, naturally, and proposals to make plen>64 + L=1 into BCP is worrisome accordingly— and providing reachability at global scope addresses to nodes on another link by proxying ND. In this case, the CE router sends inbound packets to the 6LBR which retransmits them to the destination on its 6LO interface only if it has a host route and a ND proxy binding.

In that scenario, to say that the part of the IPv6 address that follows the on-link prefix, which‚ let me repeat, is really a routing prefix and its address configuration operational semantics are completely orthogonal to this discussion, identifies the interface on the 6LBR node and not the real destination address at the downstream 6LO node's interface, is to make a technical distinction in the context of forwarding and routing without any good purpose.

For that reason, I’d prefer not to use “interface identifier” for the part of a unicast address that follows a routing prefix, particularly the part the follows the routing prefix advertised in RA message PIO elements with A=0 and L=1.


--james woodyatt <[e-mail] <mailto:[e-mail]>>