Thanks: GSS Server without secret key?

[Mike Friedman]

On Fri Nov  7 01:57:42 2003, Oliver Schoett said:

> The design seems to be asymmetric in that the need to store a secret 
> long-term key at the client has been avoided (the client only needs to 
> store its TGT), but a secret long-term key at the server is still 
> necessary.  I am afraid our customer will complain about this ...

Oliver,

Well, it's actually a little more symmetric than that.  If the client is
acting on behalf of a user at a terminal, then the secret long-term 'key'
IS stored - in the user's biological memory (in the form of a password
that gets converted to the key).  The server's keytab plays a role analogous
to a human user's memory.

If a client must authenticate while unattended by a human, then the key
WOULD have to be stored somewhere on the client.

BTW:  I'm speaking basic Kerberos protocol here, not about particularly
about GSS.

Mike

------------------------------------------------------------------------------
Mike Friedman                             System and Network Security
mikef at ack.Berkeley.EDU                    2484 Shattuck Avenue
1-510-642-1410                            University of California at Berkeley
http://ack.Berkeley.EDU/~mikef            http://security.berkeley.edu
------------------------------------------------------------------------------