: Re: authors 48 hours: RFC 4005
David Mitton <[email protected]> Mon, 30 May 2005 23:46:14 -0400
| Newsgroups | gmane.ietf.aaa |
|---|---|
| Message-ID | <[email protected]> |
Suggested changes
Dave.
------
Text changes for RFC 4005 <draft-ietf-aaa-diameter-nasreq-17.txt>
Section 9.6. RADIUS Vendor Specific Attributes
para 1: s/recommended/example/
after para 1: <add >
A system communicating between Diameter and RADIUS MAY have specific
knowledge of vendor formats, and MAY be able translate between the
two formats. However, given the deployment of many RADIUS vendor
formats that do not follow the example format in RFC 2865 [RADIUS],
(e.g. those that use a longer vendor type code) the translations in
the next two sections, will not work in general for those VSAs. RFC
2865 states that a robust implementation SHOULD support the field as
undistinguished octets.
Systems that don't have vendor format knowledge, MAY discard such
attributes not knowing a suitable translation. An alternative format is
under consideration [VSAdraft] that proposes encodings that would
preserve the native information and not require vendor knowledge in
the gateway system.
The following sections are an example for translating RADIUS VSAs
that use the example RADIUS format, and Diameter VSAs that have type
codes less than 255, and value field lengths less than 252.
<endadd>
9.6.1. Forwarding a Diameter Vendor AVP as a RADIUS VSA
Add to para 1, first sentence:
For Type codes less than 255, the value field length MUST be less
than 252 or the AVP will be discarded.
Section: 9.6.2. Forwarding a RADIUS VSA as a Diameter Vendor Specific AVP
Change from:
Diameter AVP length = length of AVP (heade
To:
Diameter AVP length = length of AVP (header + data)
Informative References (section 13.2, page 79)
<Add after [DiamMIP]>
[VSAdraft] D. Mitton "Diameter/RADIUS Vendor Specific AVP Translation",
draft-mitton-diameter-radius-vsas-00.txt, "Work in Progress", April 2005
<endadd>