[jcifs] Re: Getting jCIFS to work with NetApp

Michael B Allen mballen at erols.com
Fri Mar 22 10:27:56 EST 2002

On Thu, 21 Mar 2002 14:29:24 -0000
"Inggs, Andrew" <andrew.inggs at thus.net> wrote:

> > From: Michael B Allen 
> > 
> > Well there are quite a few differences in that Session Setup AndX but
> > none of them should matter. I think the best course of action 
> > here is to just set all the necessary jCIFS properties to make it
> > look just like smbclient.
> > 
> I've got it as close as I could and it still fails (Ethernet dumps
> attached).  These are the only differences that I can still see
> between the smbclient and jCIFS Session Setup AndX Requests:
>  - Process ID, SMB Header
>    smbclient: 578 (same as VC Number)
>    jCIFS:     33421

Doesn't matter.

>  - Multiplex ID, SMB Header
>    smbclient: 1 (same as Negotiate Protocol Request)
>    jCIFS:     2

It's possible that NetApp has some weird bug that requires this to be
1 which would explain why smbclient uses mid 1 twice. I doubt it though.

>  - Passwords
>    Although the session keys are the same, the encrypted passwords are
>    different.  This could be because different algorithms or salts are
>    used -- I don't know how the encryption works, I'm just noting that
>    they're different.

The session key could be it. It's been a long time since I had to deal
with any crypto. It's always just worked. The encrypted passwords will
always be different because it's generated based on the Challenge returned
in the Netgotiate Response which is a random number. But if NetApp
expects the session key to be factored in somehow then we have a problem.

Unfortunately I do not know how to proceed here. I could try asking
samba-technical but they have never responded to my queries in the
past. I will probably have to try to contact NetApp but that's probably
an equally futile exercise; our little Java program just doesn't have
any market share :~(


May The Source be with you.

More information about the jcifs mailing list