Configuration and Provisioning for Wireless Access Points (CAPWAP) Problem Statement
draft-ietf-capwap-problem-statement-02

Received via General Area Review Team
---------- Forwarded Message ----------
Date: 18. august 2004 13:13 +0300
From: john.loughney@nokia.com
To: gen-art@alvestrand.no
Subject: Review of draft-ietf-capwap-problem-statement-01.txt


> http://www.ietf.org/internet-drafts/draft-ietf-capwap-problem-statement-0
> 1.txt
>
>    CAPWAP Problem Statement (Informational) - 6 of 6
>    Token: Bert Wijnen
>    REVIEWER: John Loughney

Summary:

I would put a discuss, based upon my comment #3 - security considerations
section.

Comments:

1) Abstract is a bit thin, I would suggest a little more detail here.
2) Section 2 is the main problem.  I actually think that the work
  should be done, but I have to say that section 2 did not convince
  me that the work should be done.  However, I am also of the opinion
  that Problem Statement documents should be short and to the point.
  I am not sure if it is worthwhile to spend 3 more months improving
  this section
3) My biggest issue is with the security considerations.  Shouldn't this
  section discuss possible threats cause by the lack of a solution, and
  also a short discussion on what new threats could be created?  One of
  the downsides to centralization is the ability for someone to gain
  control of a large number of devices.  If the central control entity
  is compromized, your in big trouble.

NITS

1) Standard 3667-3668 text needs inserting
2) I didn't run the ID-NITs script, nor did I do a reference check.


---------- End Forwarded Message ----------

[Ballot comment]
--------- comment from AAA doctor:

>  o draft-ietf-capwap-problem-statement-01.txt
>    CAPWAP Problem Statement (Informational) - 6 of 6

This document is very good. One additional item that
could perhaps be discussed in the security considerations
section is related to problem #4, securing a distributed
set of access points. I think problem #4 implies that
people want to move some of the tasks of the APs to the
controller so that the compromise of an access point would
not compromise, say, a RADIUS shared secret or perhaps
not even some of the keys used to protect the link layer.
Security considederations does not point this out. It
does say something about looking into physical security,
but perhaps it could be expanded into that as well as
the assignment of security parameters to the different
parts of the system.

[Ballot discuss]
If the responsible AD says that the group is better served by not worrying abou this document any more, I'll drop this one.
Still..... from John Loughney's review:

Shouldn't the security considerations section discuss possible threats cause by the lack of a solution, and also a short discussion on what new threats could be created?  One of the downsides to centralization is the ability for someone to gain control of a large number of devices.  If the central control entity is compromized, your in big trouble.