[radext] Fwd: radext-ieee802 discussion at ietf-87

Bernard Aboba <bernard_aboba@hotmail.com> Tue, 30 July 2013 23:35 UTC

Message-ID: <BLU404-EAS305E35DACFD4F96BBDD333093560@phx.gbl>
Content-Type: multipart/alternative; boundary="Apple-Mail-F85330FE-5667-4D8A-9426-A552789D0D43"
Content-Transfer-Encoding: 7bit
References: <51F7E8AF.4050503@sbcglobal.net>
From: Bernard Aboba <bernard_aboba@hotmail.com>
Date: Wed, 31 Jul 2013 01:35:22 +0200
To: radext@ietf.org
MIME-Version: 1.0 (1.0)
Subject: [radext] Fwd: radext-ieee802 discussion at ietf-87
Precedence: list



Begin forwarded message:

> From: Mick Seaman <mickseaman@sbcglobal.net>
> Date: July 30, 2013, 6:24:15 PM GMT+02:00
> To: Brian Weis <bew@cisco.com>, Joe Salowey <jsalowey@cisco.com>,  Bernard Aboba <bernard_aboba@hotmail.com>
> Subject: radext-ieee802 discussion at ietf-87
> 
> Well I had hoped that this was being put to bed more or less "as is" since I think it is probably a never ending subject.
> [I make that comment in the light of the recent 'OmniRAN' exec study group in 802, whose mission seems to be to
> range between some new over-arching access architecture for both wired and wireless at one extreme and complaining
> that they cant find the necessary radius attributes or don't like the way they are written down at the other. I'm prepared to
> believe in existence of some gaps, so the effort may produce some further items. It would be good to get the base-line down].
> 
> If you are moving in the direction of being able to quote any/all EAPOL announcement items in Radius then I would suggest
> you not leave out the Organizationally Specific items. This (organizational specific extensions in general) seems to be a usefulsafety valve/play pen for certain government related orgs who both want to have 'their stuff' in 1X while at the same time not wanting to
> be very specific about what this 'stuff' is.
> 
> Seem to be three ways of going forward as philisophy:
> 
> 1. Add items from EAPOL announcements to Radius one by one if and when they are understood in Radius terms, in particular when it is known they are not best derived from existing Radius attributes.
> 
> 2. Provide a way for carrying the EAPOL announcement items in general, but state when the same info would be better carried
> in another Radius attribute of established use (and then possibly translated/copied into EAPOL announcements as a result of
> some (authenticator local) policy.
> 
> 3. Finish radext-ieee802 now by most expedient means and add additional items later as part of new work.
> 
> I would favour some mix of 1 and 3. I think something that falls between 1 and 2 but is not either of them is probably not a good idea.
> 
> 
> Brian, we could follow up with discussion in York on this topic if it is not put to bed before then.
> 
> Mick
>

[radext] Fwd: radext-ieee802 discussion at ietf-87 Bernard Aboba