Comments on IPv6 Prefix Subdelegation

[Fred Baker <[e-mail]>]

Let me make an introductory comment on:
http://tools.ietf.org/html/draft-baker-ipv6-prefix-subdelegation
  "Prefix Sub-delegation in a SOHO/SMB Environment", Fred Baker, 27- 
Jul-09,
  <draft-baker-ipv6-prefix-subdelegation-00.txt>

In IPv6 Operations, we have two posted documents right now that  
comment on prefix subdelegation. These are:

http://tools.ietf.org/html/draft-donley-ipv6-cpe-rtr-use-cases-and-reqs
  "Use Cases and Requirements for an IPv6 CPE Router", Chris Donley,  
Deepak
  Kharbanda, John Jason Brzozowski, Yiu Lee, Jason Weil, Kirk  
Erichsen, Lee
  Howard, Jean-Francois Tremblay, 2-Jul-09,
  <draft-donley-ipv6-cpe-rtr-use-cases-and-reqs-00.txt>

http://tools.ietf.org/html/draft-ietf-v6ops-ipv6-cpe-router
  "IPv6 CPE Router Recommendations", Hemant Singh, Wes Beebee, 25- 
Mar-09,
  <draft-ietf-v6ops-ipv6-cpe-router-00.txt>

The premise is that an ISP might delegate a PA prefix to a SOHO/SMB  
network, perhaps using DHCP or etc. It would be nice if the prefix  
could be in turn sliced into /64 prefixes and sub-delegated to the  
various LANs in the subsidiary network.

draft-ietf-v6ops-ipv6-cpe-router is trying to recommend to vendors  
that they should build CPE routers in a certain way, and specifies in  
part how sub-delegation would work. In my opinion as WG chair, I would  
rather that it said "do RFC X" than "do the following algorithm", as  
one might want to change the algorithm and the proposed algorithm has  
not been proven operationally. In general, I would like 6man to take  
on the work of describing that algorithm.

I threw draft-baker-ipv6-prefix-subdelegation together very quickly  
for the purpose of saying "I would want you to reference something  
like <this>". That said, it is at least a first step, and may be the  
right answer for the moment. I would appreciate it if 6man could take  
a look at the discussion on sub-delegation in the two CPE drafts and  
at this draft, and decide first whether the draft is a reasonable  
first step toward solving the problem that the CPE drafts target, and  
then further decide whether and with what authors they would like to  
finish that discussion. I'm throwing no personal ego in here - if  
someone else would like to respond to the question, less work on my  
part sounds good to me.

Your opinions, please...