Archive

>
>On Sat, Aug 20, 2005 at 12:58:41AM -0500, Nicolas Williams wrote:
>> On Fri, Aug 19, 2005 at 11:10:35PM -0400, Shumon Huque wrote:
>> > I think the Kerberos community needs to work on standardizing password 
>> > based pre-authentication mechanisms invulnerable to dictionary attack 
>> > (perhaps EKE, AEKE, SPEKE, SRP etc). Hardware pre-authentication
>> > mitigates the threat somewhat. But PKINIT isn't really an option for 
>> > the many sites that don't plan to authenticate users with public key 
>> > credentials (or deploy PKI).
>> 
>> Did you attend the SACRED WG meeting at IETF 63?
>
>No, but I just read a summary of their session. The credential
>mobility stuff isn't really relevant to (non PKINIT) Kerberos
>sites. So I assume you are referring to the IPR issues with those
>protocols. It sounds like there might be some hope though!
>
>> > At one time, some of us were talking about an EAP method that
>> > transported Kerberos messages directly. It seems to me that putting
>> > some effort into completing that work would be immediately useful
>> > to Kerberos sites that need to deploy 802.1X or 802.11i soon.
>> 
>> Indeed.  That would be an example of framework bindings of a security
>> mechanisms, as opposed to EAP-GSS, which would be a bridge.
>
>Agreed.
>
>Some folks at Penn are interested in working on an EAP-Kerberos 
>method. Is there anyone else interested in this?

Hi Shumon, 

I`ve also thought about a new EAP method which makes use of Kerberos.
When designing a new EAP method, it must extend current available set
of EAP methods by a new mechanism.

Some weeks ago, I`ve asked publicly in the Kerberos Mailing List what
the folks expect from a EAP-Kerberos method. See http://mailman.mit.edu/pipermail/kerberos/2005-July/008131.html
Some opinions can be found there.

There already exists a Kerberos extension to TLS, RFC 2712 (Oct.99),
which can be run in EAP-TLS, so the question is: 

* Is there need for EAP-Kerberos at all? * 

So before all, we should investigate in how far EAP-Kerberos improves
the TLS-based solution. 

For instance, the mandatory resistance to dictionary attacks. Thomas Wu
has given in his Kerberos paper a hint how to mitigate this, however,
even if there is a strong-password protocol without IPR claims,
strong password methods suffer in general from heavy computation and thus
the EAP method would have worse performance.

Any comments are welcome
Thomas

_______________________________________________
SECMECH mailing list
SECMECH@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/secmech

AW: Re: [SECMECH] Framework Bindings Vs. Mechanism Bridges