[OpenAFS] Re: Authentication problems with 1.0.4 (Charles Clancy)

[Christian Pretz]

>> > I can configure the client, but whenever I try to log in, it always

>> > says Error: 62 (password is incorrect)

>> sorry to say, but i built that release and never testing standard the

>> standard klog.  we use aklog here (even on windows nt).  i assumed
that
>> since everything else worked, everything was fine.

>> i get a variety of strange messages from klog from 'badly skewed
clocks'
>> to 'missing realm' -- the whole windows nt klog is strange since it
>> does krb v4 to authenticate you.

>If you run command-line klog, you'll see:
>C:\Documents and Settings\mgrtcc>klog mgrtcc
>Password:
>** Can't determine local cell name!
>Unable to authenticate to AFS because password was incorrect.

>This is a bit more informative.  Where does klog try and find the
ThisCell
>information?

>If I change my clock to the wrong time, I see:
>C:\Documents and Settings\mgrtcc>klog mgrtcc
>Password:
>** Can't determine local cell name!
>Unable to authenticate to AFS because server and client clocks are
badly
>skewed.

>I find that interesting.  If it can't determine the local cell name,
how
>does it know which server to check the time against?

I had the same errors when I tried to set up the 1.0.4 Client for
Windows NT.
In my case the problem was caused by the afsdcell.ini. The
afsdcell.ini-file which was created by the afsconfig-tool does not work.
(-> ** Can't determine local cell name!)
Since I am using the afsdcell.ini of my working Transarc AFS 3.6  on
OpenAFS 1.0.4, my authentication problems have been solved.
Seems to me that there might be a problem in the created
afsdcell.ini-file. Have not found out yet what it was.

Greetings
 Chris

--
+--------------------------------------------------------------------------=
-----+

  No trees were killed in the sending of this message. However a large
  number of electrons were terribly inconvenienced.
+--------------------------------------------------------------------------=
-----+

  Christian Pretz
  Leibniz-Rechenzentrum                  Tel.:(089) 289-28770
  Barer Str. 21                          E-Mail:
Christian.Pretz@lrz-muenchen.de
  D-80333 M=FCnchen
+--------------------------------------------------------------------------=
-----+

             Lubarsky=B4s Law of Cybernetic Entomology:
                  There is always one more bug.
+--------------------------------------------------------------------------=
-----+