draft-ietf-rap-rsvp-authsession-05.txt

Louis-Nicolas Hamer <[email protected]> Sun, 10 Nov 2002 21:01:13 -0500
Newsgroups gmane.ietf.rap
Message-ID <[email protected]>
  Hi all,

As Bert has indicated, a set of comments from Eric Rescorla had not been 
addressed in the last revision.
The comments were related to the security aspects of the draft. A few 
modifications were made
to ensure the draft specified in more details some security aspects.
Summary of ALL changes:

-1: DNS Spoofing: Eric identified a DNS Spoofing issue. Because of this 
flaw,
a few fields were removed from the SOURCE_ADDR and DEST_ADDR S-TYPEs, 
specifically,
the FQDN, ASCII_DN & UNICODE_DN.

-2: Key rollovers: The example for shared symmetric keys was missing one
field, the AUTH_ENT_ID (I added that to the example). I also added the 
following clarification:
"Since multiple keys may be configured for a particular
   AUTH_ENT_ID value, the first 32 bits of the AUTH_DATA field MUST
   be a key ID to be used to identify the appropriate key.

-3: Time synch: Added a sentence to discuss why it is important.

-4: Changed "should" to "SHOULD" in the sentence: "Triple-DES encryption 
is supported in many Kerberos implementations
   (although not specified in [RFC-1510]), and SHOULD be used over
   single DES."

-5: PGP section: wrong terminology was used - It was removed.

-6: X.509 V3 section:
Clarified the certs and crls. Changed the X509_V3_CERT field to be a DN.

-7: Kerberos.  Added the clarifications needed
about the client and server:
" In this request, the client
   (router/PDP) sends (in cleartext) its own identity and the identity
   of the server (the authorizing entity taken from the AUTH_ENT_ID field)
   for which it is requesting credentials .

-8: Clarifications added to section 6.4.
Clarifications added about the danger to rely upon an insecure database 
(such
   as DNS or a public LDAP directory).

Document available @
http://www.ietf.org/internet-drafts/draft-ietf-rap-rsvp-authsession-05.txt

Thanks to Bert for his helpfull assistance. And many thanks to Eric for 
his comments & suggestions.
The draft has been re-inputted into the RFC-Editor's queue.

Cheers,
L-N