Re: [v6ops] I-D Action: draft-ietf-v6ops-unique-ipv6-prefix-per-host-07.txt

[DY Kim <[e-mail]>]

in line...

—-
DY


> On 4 Aug 2017, at 22:58, David Farmer <[e-mail]> wrote:
> 
> I don't see how whole subnet or prefix assigned to just a single host is incompatible with with RFC4291 or RFC4291bis in any way.

Subnet, to my understanding, is a shared network comprising of a collection of hosts managed (bordered) by one or more routers. A subnet prefix is one assigned to the whole subnet wherein all hosts share the same prefix.

When each host in such a share network (subnet) is assigned a unique prefix (and hosts don’t share the same prefix), that prefix is not a subnet prefix and should be name differently, say, ‘host prefix’ or ‘node prefix’.

RFC4291bis defines (comments) only the subnet prefix. The draft of this thread title deals with a prefix type not defined (commented) in the RFC4291bis (to-be-)standard. In that, the draft under discussion in this thread could be said ‘not in compliance with’ the RFC4291bis.

This is my logic. Odd yet?


> That said there was some discussion of adding a reference to RFC7934 - Host Address Availability Recommendations, which introduces the concept of assigning a whole /64 to an individual host.  I think its still think that is a good idea, and maybe adding a reference to this would be a good idea too. It does represent a bit of an evolution in the subnet model.