eap Mailing List: Re: hopefully final changes for draft-ietf-eap-keying

[Yoshihiro Ohba (yohba]

It seems that the change came from the following IESG DISCUSS comment:

https://datatracker.ietf.org/idtracker/draft-ietf-eap-keying/comment/73693/?

"
>An EAP authenticator MUST NOT share any keying material with
>another EAP authenticator, since if one EAP authenticator were
>compromised, this would enable the compromise of keying material on
>another authenticator. 

This text needs to be fixed to allow 802.11r style key hierarchies.
Previous text gets this correct.
"

As far as I understand 802.11r has a distributed authenticator model
in which an authenticator may span multiple APs.  Hence I don't think
this implies key sharing among multiple authenticators.

Regards,
Yoshihiro Ohba



On Fri, Nov 16, 2007 at 04:48:46PM -0800, Bernard_Aboba [at] hotmail.com wrote:
> Forwarding for Dan.   
> 
> 
> --------------------------------------------------------------------------------
> From: Dan Simon
> Sent: Friday, November 16, 2007 9:10 AM
> To: eap [at] frascone.com
> Cc: jari.arkko [at] piuha.net; iesg [at] ietf.org
> Subject: Re: [eap] hopefully final changes for draft-ietf-eap-keying
> 
> 
> Having looked over the proposed changes, I share the concern about the 
> modified language relating to key-sharing.  The new language is disturbingly 
> vague-would any key-sharing among authenticators or EAP servers be 
> acceptable, as long as the shared keys have passed through a key derivation 
> step?  Since virtually all the keys in the hierarchy were derived at some 
> point, the restriction on key-sharing would effectively disappear completely 
> under the new language.
> 
> I would propose more restrictive changes to the language, but after looking 
> at the documents of the HOKEY group (whence, as I understand it, the 
> initiative to loosen the key-sharing language emerged), I can't for the life 
> of me see any incompatibilities between their work and the current language.  
> Perhaps someone could articulate a clear, plausible example of a case where 
> adherence to the current language significantly impedes HOKEY progress-or for 
> that matter, progress on the design of any other useful protocol standard?
> 
> So my vote would be to leave the key-sharing portion of the text from -18 
> intact.
> 
>                                 Dan Simon

> _________________________________________________________________
> To unsubscribe or modify your subscription options, please visit:
> http://lists.frascone.com/mailman/listinfo/eap
> 
> Arhives: http://lists.frascone.com/pipermail/eap