I’d like to setup an active directory. Hardware, 2 identical systems P4’s that will serve as my AD root, files and folders, and 3 printers. Object, I’d like to set it up so that both the 2 P4’s will be running as AD root – so if one fails, the other one will run while I restore the other one. Could I use clustering for my AD structure? Is it worth it? Is there other ways for me to setup if I wanted 2 machines replicating user accounts, Printer queues, and files and folders?

AD is multimaster. You do not need clustering. Make both DC's Global Catalogs. "Christmas and the New Year are actually two holidays.... I suppose you could say 'Merry Christmas' and 'Happy New Year,' but you probably have sh!t to do." --Jon Stewart, on saying "Happy Holidays"

If I make both of them as DC's - will both authenticate the user's log in? And will both servers have the files and folders for user access? If a user access and modify's a file, will both of the servers have the updated modified files or will it replicate?

File storage has nothing to do with DC's. That is a file server role. Yes, both can authenticate the user. Please help survivors of Hurricane Katrina. www.redcross.org

I would suggest using the 2 servers as AD one being a global catalog with most of the master roles on one and the infrastructure master on the second box. The infrastructure master should not be in a global catalog server, microsoft recommendation. Also take a look at DFS (distributed files system) in windows 2003 server and set up replication so the files will be replicated across the servers. In my experience the easiest thing for printer server redundancy is to install the printers onto both servers (network printers i am assuming). Than on the client side install both printers from each server and the users can change over if a server goes down manually. A lot easier to setup than clustering and less resource intensive on the server hardware.

"I would suggest using the 2 servers as AD one being a global catalog with most of the master roles on one and the infrastructure master on the second box. The infrastructure master should not be in a global catalog server, microsoft recommendation." Per Microsoft recommendation, infrastructure masters should not be GC's with two exceptions: A. All DC's are GC's. If the user has two DC's, with both as GC's, he meets this exception. B. There is only one domain in the forest. This user has only one domain. He also meets this exception. http://support.microsoft.com/kb/223346/EN-US/ Without the second as a GC, the user does not have full redundancy. Therefore, both should be GC's in this situation. As for FSMO role holders, the fastest of the two DC's should be the PDC and RID master. The other should be the schema, infrastructure, and domain naming master. However, the user should be aware of issues with the infrastructure master as a GC issue. Should the user ever move to a multidomain environment, he must make sure all his DC's are GC's, or move the infrastructure master role to a DC that isn't a GC. Please help survivors of Hurricane Katrina. www.redcross.org

Thanks for all the great reconmendation / information... I will read up on it and see how I can put it into our AD structure. Thanks again.