Re: [AAA-DOCTORS] Functionality in draft-ietf-dime-mip6-integrated vs. draft-ietf-mip6-radius

Hi Pasi, 

my intention was to align it with draft-ietf-dime-mip6-integrated as
much as possible. 

The reason why it is different is simple: There are folks interested to
implement draft-ietf-dime-mip6-integrated and hence we have pressure to
finish it asap. I will work on the alignment once
draft-ietf-dime-mip6-integrated is finished. 

Ciao
Hannes

>-----Original Message-----
>From: Pasi.Eronen@nokia.com [mailto:Pasi.Eronen@nokia.com] 
>Sent: 02 December, 2008 13:39
>To: avi@bridgewatersystems.com; 
>kchowdhury@starentnetworks.com; Tschofenig, Hannes (NSN - 
>FI/Espoo); jouni.nospam@gmail.com; Korhonen, Jouni (NSN - 
>FI/Espoo); charliep@wichorus.com; julien.bournelle@orange-ftgroup.com
>Cc: dave@frascone.com; dromasca@avaya.com; 
>jari.arkko@piuha.net; marcelo@it.uc3m.es; 
>julien.laganier.ietf@googlemail.com; aaa-doctors@ietf.org
>Subject: Functionality in draft-ietf-dime-mip6-integrated vs. 
>draft-ietf-mip6-radius
>
>
>Jouni, Avi, and others,
>
>I've been re-reading the other drafts related to 
>draft-ietf-dime-mip6- integrated, and I have a question which 
>may be quite important the issue at hand: do you expect 
>draft-ietf-dime-mip6-integrated to provide the same 
>functionality/semantics as the "integrated scenario" parts of 
>draft-ietf-mip6-radius?
>
>So far, I had assumed the intent was to provide identical 
>functionality:
>perhaps using different AVP/attribute formats, but a 
>translation agent could perform lossless translation in either 
>direction.
>
>However, on further reading, it seems the current versions of 
>mip6-radius and dime-mip6-integrated do different things (and 
>neither is a subset of the other). For example:
>
>- The DIME draft allows sending arbitrary number of home link
>  prefixes (zero or more) in both request and response messages;
>  the MEXT draft allows just zero or one.
>
>- The DIME draft allows sending either the FQDN, or IP address(es),
>  or both for a single HA (and a message can contain information
>  about more than one HA). The MEXT draft allows sending multiple
>  FQDNs and/or IP addresses, but there's no way to say that
>  some FQDN and IP address(es) are for a single HA.
>
>- The MEXT draft allows the AAA server to assign a home address
>  to the MN; the DIME draft does not support this (both support
>  communicating the home link prefix, though).
>
>- The DIME draft allows sending the "realm" associated with
>  a home agent; the MEXT draft doesn't.
>
>- The DIME draft allows sending the IPv4 address of a MIPv6
>  home agent that supports draft-ietf-mext-nemo-v4traversal;
>  the MEXT draft does not.
>
>- The MEXT draft supports DNS updates; the DIME one does not.
>
>(BTW -- all this is about the "integrated scenario"; split 
>scenario is a topic for a different day...)
>
>If these differences in functionality/semantics are 
>intentional, and there's no desire to allow lossless 
>translation in either direction (even assuming new code in the 
>RADIUS/Diameter translation agent), then using identical AVP 
>formats/numbers isn't that big deal.
>
>So, I'm wondering whether these differences are intentional, 
>or perhaps artifacts of not-totally-synchronized work in two WGs?
>
>Best regards,
>Pasi
>
_______________________________________________
AAA-DOCTORS mailing list
AAA-DOCTORS@ietf.org
https://www.ietf.org/mailman/listinfo/aaa-doctors