Duplicate machine GUIDs

[Bruce]

Hi,

We've just migrated to Win2K Server/WinXP. Don't ask me how it
happened but we have ended up with duplicate machine GUIDs on our
network. For each XP PC there are maybe 2 or 3 others with the same
GUID.

[For Info: It happend because we used the same bunch of network cards
for RIS-ing the PCs. The onboard network cards weren't compatible with
RIS so we swapped network cards in and out as we went along. I don't
know if anyone else has experienced this?]

Anyway, the end result is that we now have duplicate machine GUIDs in
our AD. We haven't noticed any problems yet. Does anyone know if
it will cause any problems in the future? Also, how do we solve the
problem so that all PCs have a unique GUID?

Thanks,

Bruce.

 

[Herb Martin]

[For Info: It happend because we used the same bunch of network cards

for RIS-ing the PCs. The onboard network cards weren't compatible with
RIS so we swapped network cards in and out as we went along. I don't
know if anyone else has experienced this?]

Still shouldn't have happened as the GUIDs should be chosen based
on that AND also Time with a Random number factor.

If it did happen that way then the algorythm is defective.

Are we talking GUID or SID?

 

[Bruce]

[For Info: It happend because we used the same bunch of network cards
for RIS-ing the PCs. The onboard network cards weren't compatible with
RIS so we swapped network cards in and out as we went along. I don't
know if anyone else has experienced this?]

Still shouldn't have happened as the GUIDs should be chosen based
on that AND also Time with a Random number factor.

If it did happen that way then the algorythm is defective.

Are we talking GUID or SID?

The GUIDs of the machines in the AD seem to be an exact copy of the
MAC address of the NIC that was used to RIS them with (but padded with
zeros at the beginning).

When tried to RIS a machine with a NIC that had been used previously
(to RIS) it warned us that there was a duplicate GUID and suggested
going ahead with the RIS anyway and sorting out the duplciate GUIDs
later, but didn't provide any more information on how to do that.

My Win2K knowledge is a little patchy, I had assumed the SIDs (used in
WinNT4) had been replaced by GUIDs (in Win2K). If there is still a SID
per machine then why do we also have a GUID per machine (we affectivly
have 2 unique identifiers per machine)?

Thanks,

Bruce.

 

[Herb Martin]

My Win2K knowledge is a little patchy, I had assumed the SIDs (used in

WinNT4) had been replaced by GUIDs (in Win2K). If there is still a SID
per machine then why do we also have a GUID per machine (we affectivly
have 2 unique identifiers per machine)?

No, they have been supplemented (not replaced.)

--
Herb Martin

[For Info: It happend because we used the same bunch of network cards
for RIS-ing the PCs. The onboard network cards weren't compatible with
RIS so we swapped network cards in and out as we went along. I don't
know if anyone else has experienced this?]

Still shouldn't have happened as the GUIDs should be chosen based
on that AND also Time with a Random number factor.

If it did happen that way then the algorythm is defective.

Are we talking GUID or SID?

The GUIDs of the machines in the AD seem to be an exact copy of the
MAC address of the NIC that was used to RIS them with (but padded with
zeros at the beginning).

When tried to RIS a machine with a NIC that had been used previously
(to RIS) it warned us that there was a duplicate GUID and suggested
going ahead with the RIS anyway and sorting out the duplciate GUIDs
later, but didn't provide any more information on how to do that.

My Win2K knowledge is a little patchy, I had assumed the SIDs (used in
WinNT4) had been replaced by GUIDs (in Win2K). If there is still a SID
per machine then why do we also have a GUID per machine (we affectivly
have 2 unique identifiers per machine)?

Thanks,

Bruce.[/QUOTE]