ntlm_auth - rlm_mschap: No User-Password configured. Cannot create NT-Password.

[Phil Mayers]

Peter de Groot wrote:
> 
> Eeeerrk....
> 
> Given that I suspect that this form of authentication  is going to be 
> popular ... I was wondering if
> it could be "broken out" into a test program to add to the others....

See the "eapol_test" program from the "wpa_supplicant" package. Works 
very well.

Alternatively, since the response for a given username/password and 
challenge is always the same, you can just run one auth, record those, 
and re-use them. The obvious insecurity behind this is why the RPC call 
used to pass username/challenge/response must be authenticated with 
domain machine account credentials.