Draft:  draft-ietf-mip6-precfgkbm-03.txt
Reviewer: Spencer Dawkins [spencer@mcsr-labs.org]
Review Date: Wednesday 7/6/2005 8:33 AM CST
Telechat Date: July 7, 2005

Summary: this document is very close to being ready for publication 
as Proposed Standard, with some questions.

Review comments/questions:
--------------------------
I agree with Jari's suggestion for a title change, to ""Securing 
Mobile IPv6 Route Optimization Using a Shared Key".

It was not clear to me what happens if a mobile node is trying to use 
the mechanism described in this draft, but does not actually share a 
key with the correspondent node. Should the draft recommend falling 
back to return routability? or should communication simply fail until 
the two nodes are correctly configured, so that any humans in the loop 
see that the intended optimization is not taking place? At the very 
least, I would expect to see at least these choices mentioned 
explicitly.

The draft calls out "correspondent node must trust the mobile node to 
behave" in a couple of places. If this really is a significant 
consideration, I wonder about checking return routability in addition 
to the shared key mechanism (you could still get the low latency when 
beginning communication, but would also notice misbehaving mobile 
nodes eventually).

There are only two references, one normative and one informative. I'm 
not sure whether we are splitting the reference section in all drafts 
these days or not (just wanted to mention it).

I'd also like to say that the draft is pretty clear to the non-MIP6 
reader - thank you!