[jcifs] NTLM usrname/password failure after each 5 mins

[Michael B Allen]

Alexandr,

Honestly I can't look into this much. I'm busy trying to start a company
(that has nothing to do with jcifs BTW). I can answer emails but I'm
afraid the only way you'll really get to the bottom of the problem is to
look at the jCIFS source and figure out how it works for yourself. That's
the nature of Open Source sometimes.

As for what I think is happening, I don't know. My understanding of the
code was that a transport would only close if the connection was idle. So
if you are using a fixed number of credentials that are all initialized
on startup and there is no further communication with the DC then the
transport closing is normal and we're back to the original response which
is that it's a problem inherent to the NTLM multistep handshake. Note
however that it will likely not occur in production because the chances
of someone logging in at the exact moment the transport closes are slim
(with a sufficiently high soTImeout).

Mike

On Fri, 21 Jul 2006 10:11:44 +0400
"Alexandr Podoplelov" <alepod at gmail.com> wrote:

> Yes, it's a stress test and I do use different credentials. The problem
> occured after the first read timeout. Can it be possible to increase the
> timeout value. I use jcifs of version 1.2.9 . Maybe it makes sence to add
> some extra log prints into sources?
> 
> //Alexander
> 
> 2006/7/20, Michael B Allen <mba2000 at ioplex.com>:
> >
> > Actually I forgot you're stess testing. Are you using different
> > credentials? That would explain the different sessions.
> > --
> > Michael B Allen
> > PHP Extension for SSO w/ Windows Group Authorization
> > http://www.ioplex.com/
> >
> 


-- 
Michael B Allen
PHP Extension for SSO w/ Windows Group Authorization
http://www.ioplex.com/