Hi bash3r,
Quote: You did no kind of migrate or upgrading correct, just new box and W2K3 added to your existing network?
It was a new box added to the exisiting domain. Since we wanted minimal disruption to our network we did bring in a loaner NT4.0 BDC (not originally on our network)and upgrade it to PDC then to W2K3 since the PDC should be the first to be upgraded to W2K3.
Quote: Your W2K3 server is now your global catalog server
Our W2K3 is now our global catalog server, DNS server, and WINS server.
Quote: did you have to make changes to any of your NT servers once the W2K3 server was online?
Because we'd only used WINS before we had to configure DNS on all our servers and clients. We also had to change our Primary WINS server address to point to the new server.
Quote: Your domain name stayed the same
Yes, it doesn't need to change.
Quote: Did you change the hostname of the W2K3 server different from the failing NT PDC server?
Yes, only because we wanted to keep our network up until it was time to move the W2K3 server into place. Our actual down time was less than 8 hours including data transfer. We used Snapshot for 26GB of data.
One gotcha, if you don't alredy know, W2K3 doesn't seem to like NETBUI, at least not in hybrid mode. We removed it from all servers and clients and use only TCP/IP.
Move all your data files to a usrshr directory and your drive mappings can remain the same (i.e. E:\ drive on your old server can still be mapped as E:\ on your W2K3 server). If you map your drives through your logon script you just have to change the name of your server there. We went from HQ to HQ2 for server names. You will only have to remap the Win9x clients manually and UNC file paths may need to be reconfigured depending on your set up.
We're having a minor problem with our Exchange 5.5 Internet Mail Service not wanting to restart immediately after a system hard boot. There's no problem if the DNS server hasn't been shutdown and restarted but otherwise it takes 35 -40 minutes before the IMS will start. However, we believe this to be an issue with our spam filter which uses localhost 127.0.0.1.
Cheers.