FreeTDS Development Group

[Frediano Ziglio <freddy77 AT gmail.com>]

2010/11/10 Samuel Bayer <sam AT mitre.org>:
> Peter C. Norton wrote:
>
>> [I'm far from an expert, so apologies in advance if I'm mis-reading
>> your environment and the errors - we don't use NTLM with SQL Server]
>>
>> Are you sure that you're negotiating NTLM and not kerberos in this
>> case?  I always thought that the "trusted domain" vs. "untrusted
>> domain" indicated AD/krb5 auth.  I also thought that "windows
>> authentication" indicated the same.
>>
>> We've noticed that SQL Server 2008 can be persnickity with various
>> forms of trust in general...  Can you try the same test against a
>> win2k3 server to isolate that?
>
> I'll check with my admins, but I'm virtually certain that this really is
> NTLMv2. As I said earlier, the jTDS JDBC connection works with the useNTLMv2
> flag set, and I'm assuming that that code is correct. But I could be wrong.
>
> Unfortunately, I don't believe I have access to a different version of SQL
> Server.
>
> Does anyone know which of the precise cases described in Appendix C of
> http://davenport.sourceforge.net/ntlm.html my case corresponds to? I've been
> assuming it's either "NTLMv2 Authentication; NTLM1 Signing and Sealing Using
> the 40-bit NTLMv2 User Session Key" or "NTLMv2 Authentication; NTLM2 Signing
> and Sealing Using the 56-bit NTLMv2 User Session Key", but all this NTLM2 vs
> NTLMv2 stuff is making my head swim a bit.
>
> Thanks in advance -
>
> Sam Bayer
> The MITRE Corporation
> sam AT mitre.org

Just committed a patch for NTLMv2 and 64-bit systems due to different
structure alignment.

freddy77