I'm sorry, but this is not related to DANE (nor was it last time you came 
shopping around).

I (of course) have no problem with folk reviewing it, just making it clear that 
it isn't a DANE thing...

W

Warren Kumari
------
Please excuse typing, etc -- This was sent from a device with a tiny keyboard.

> On Sep 27, 2013, at 6:13 PM, "Hosnieh Rafiee" <[email protected]> wrote:
> 
> Hello,
> 
> I am looking for reviewers to review the
> http://tools.ietf.org/html/draft-rafiee-intarea-cga-tsig draft. It deals
> with the use of CGA (RFC 3972) or SSAS for authentication purposes during
> various types of scenarios. This is truly possible when the authentication
> is based on the source IP address. Everything in this authentication is done
> "on the fly" and only once is there a need to configure the Servers.
> Advantages:
> - If the shared secret is compromised (which is true in TSIG, you won't need
> to repeat the process of re-sharing the shared secret among the many hosts
> that used the compromised shared secret)
> - If for any reason the IP address of any nodes involved in authentication
> is changed, then the other nodes will  still be able to verify that it is
> the same node
> - It is both easier and preferable to use this approach to prove the address
> ownership while at the same time performing source IP address
> authentication.
> - The nodes don't need to go through the chain of trusts  because the magic
> of CGA or SSAS works well. This means that if the other nodes are aware of
> one node's real IP address then that is all the information the other nodes
> need  to identify this node. No matter how your node changes its IP address.
> This draft explains how to easily handle this situation, automatically,
> without the need for human intervention. 
> I am sure you will find it interesting. 
> 
> The different scenarios that are considered are:
> - Authentication during FQDN
> - Authentication during zone transfer
> - Authentication of the resolvers in stub resolvers
> - Authentication of root DNS servers to recursive DNS servers
> 
> We also considered the scenario where the server doesn't support CGA or
> SeND. The interesting thing about this draft is that you can use a CGA
> script generator that I will provide and make available to everybody. Then
> just manually set the IP address for the node that wants to use CGA-TSIG.
> The CGA-TSIG implementation will thereafter  processes the handling for all
> the CGA parameters. 
> 
> This scenario is good for the future of the Internet where IPv6 is used. I
> guess many DNS servers now support IPv6 as well as IPv4. 
> 
> I am looking forward to receiving your comments. 
> 
> Thank you,
> Hosnieh
> P.S. So if you'll review mine i'll review yours.. Deal??? :-)
> 
> _______________________________________________
> dane mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/dane
> 
_______________________________________________
perpass mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/perpass

Reply via email to