Re: [pcp] Comparison of PCP authentication

[Alper Yegin <[e-mail]>]

On Aug 7, 2012, at 9:01 PM, Sam Hartman wrote:

> If you are concerned about server DOS issues, the general solution is to
> have the server send the client an encrypted cookie so the server
> maintains no state.
> See for example RFc 6113  section 5.2.
> That's a Kerberos mechanism but it has similar properties.
> 
> EAP protocols have generally not done that, but it's certainly possibly
> to do with EAP-like things. We designed draft-ietf-abfab-gss-eap to
> permit extending to this use in the future if needed. The Moonshot
> implementation of draft-ietf-abfab-gss-eap supports the necessary
> mechanisms for no server state, demonstrating that it is possible.
> 
> Everything I say here can be applied to PANA with a bit of creativity.

PANA itself already has that property.

> _______________________________________________
> pcp mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/pcp