Hi, I am new to kerberos and would appreciate some help understanding some
basics:

What is the purpose of having a server public/private key architecture?  I
mean, when a user needs to be authenticated, the following is quite
sufficient [or is it :)]:

1.  UserID passed in plain-text to server;
2.  Server submits an encrypted "challenge"-plus-unique-session_id with the
user's password back to client;
3.  Client decrypts challenge from server with password and conducts
pre-defined scrambling (not encrypting) of plain-text;
4.  Client encrypts scrambled plain-text with unique session_id and sends
back to server;
5.  Server decrypts with previously sent unique session_id and confirms
correctness of scrambled challenge.  If ok, client authenticated and new
session_id passed for rest of the client's operations.

Each client is given its own unique session_id and the server knows which
client it is by the session_id.

Can someone please help me understand why we then need server private and
public keys (and why they have to travel as part of the authenticator)?

Many thanks

JM


________________________________________________
Kerberos mailing list           [EMAIL PROTECTED]
https://mailman.mit.edu/mailman/listinfo/kerberos

Reply via email to