Have you started using Dfs (Distrubuted File System) that has been released as a useful feature since Windows Server 2000? If you haven't, you really really should start playing around with it. It's so useful, not only because it releases the pain for managing all your network shared folders across the whole network, but also it provides some kind of fault-tolerant feature and fail-over function on your network shared resources, which saying that if Dfs is configured properly in your system, you don't have to worry about what I should do if one of the servers gets crashed or dead because the replicated network resources in Dfs are still available transparently to the end users.
I had one server failed last week, but with Dfs configured on that server, I managed to get all people back to work after only a really short down time period. So I thought I should put emphasis on it and share with other people how good this feature is. Here are some benefits you can get from Dfs if the bad thing happens on your system, like the motherboard on the server is dead:
Minimum down time
With multiple replica configured in Dfs, one failure server doesn't fall down the whole resources, the end users are still able to access the resources from other servers. People might not even notice it.
Give more time to work on the failure server
When the server gets down, people like me would feel very stressful because none of the bosses would like to have the server down for period of time, sooner gets the server back the better.
No need to update the login script any more
If you use the traditional mapping method and you changed the server after the server crash, you need to change the login script and ask all users to logoff and log back in to get the updated mapped driver. Now with Dfs, people don't have to do that, it all does automatically and transparently because all mappings are not based on the server name but the domain name now.
Windows 2003 DFS (Distributed File System)