[cifs-protocol] [REG: 111111582582159] highest_usn and tmp_highest_usn in highwatermark < max usn sent in getcchanges

[Edgar Olougouna]

[Dochelp to bcc]
[Adding case number in subject]

Matthieu,
I will investigate this and follow-up.

Thanks,
Edgar

-----Original Message-----
From: Sebastian Canevari 
Sent: Tuesday, November 15, 2011 4:55 PM
To: mat at samba.org; Interoperability Documentation Help; cifs-protocol at samba.org; pfif at tridgell.net
Subject: RE: highest_usn and tmp_highest_usn in highwatermark < max usn sent in getcchanges

Hi Matthieu,

Thanks for your question.

Someone from our team will contact you shortly.

Regards,

Sebastian

Sebastian Canevari | Escalation Engineer | US-CSS Developer Support Core (DSC) Protocol Team P +1 469 775 7849 One Microsoft Way, 98052, Redmond, WA, USA http://support.microsoft.com


-----Original Message-----
From: Matthieu Patou [mailto:mat at samba.org] 
Sent: Tuesday, November 15, 2011 4:35 PM
To: Interoperability Documentation Help; cifs-protocol at samba.org; pfif at tridgell.net
Subject: highest_usn and tmp_highest_usn in highwatermark < max usn sent in getcchanges

Hello Dochelp,


In my quite old provision it turned out that the instanceType for the root DN of the schema partition (CN=Schema,CN=Configuration,DC= ...) had a wrong value (old provision, bad knowledge of all AD stuff, so we made what we thought was good).

In a getncChanges reply from a Windows 2003R2 server the highest_usn and tmp_highest_usn are lesser than the USN of one change.

As a result samba keeps on asking this change, here is the dump of the getncchange reply.

Can you explain why the highwatermark's USNs are lower than the highest usn in the changes transmitted ?

Matthieu.

--
Matthieu Patou
Samba Team
http://samba.org