[secdir] secdir review of draft-ietf-homenet-arch-10

[Samuel Weiler <[e-mail]>]

I have reviewed this document as part of the security directorate's
ongoing effort to review all IETF documents being processed by the
IESG.  These comments were written primarily for the benefit of the
security area directors.  Document editors and WG chairs should treat
these comments just like any other last call comments.


Even though this is an architecture document, I think it deserves some 
AD attention.


First: it seems odd to set an expectation for (near) 
zero-configuration:

    It is not practical to expect home users to configure their networks.
    Thus the assumption of this document is that the homenet is as far  as
    possible self-organising and self-configuring, i.e. it should
    function without pro-active management by the residential user.

and then decline to give any guidance about, arguably, the most 
important security choice in the document:

    This document takes no position on [whether 'default allow' or
    'default deny'] mode is the default, but assumes
    the choice for the homenet to use either mode would be available.


I found the "Name spaces" section (3.7.3) a bit confusing, in part 
because it doesn't specifically name DNS at the start of the section, 
even though the details below clearly point to DNS (IDNA, possible 
conflicts with dotless TLDs).  Perhaps the second paragraph of 3.7.4, 
explaining that there are some non-DNS alternatives under 
consideration, should be moved up?  Furthermore, there are some 
particular assertions in both 3.7.3 and 3.7.4 that need to be 
reconsidered:

-- "When DNS is used as the homenet name service, it includes both a
    resolving service and an authoritative service."  Does it
    necessarily?

-- "The name space(s) should be served authoritatively by the
    homenet..."   Why is that necessary?  (Indeed, there is text in
    3.7.4 that seems to conflict with this.)

-- There is a recommendation to support DNSSEC on the authoritative
    server side (in 3.7.4).  Shouldn't there be a similar
    recommendation on the resolver side?