Using DFS for server migration
If you are an administrator in a Windows-based environment, you undoubtedly have been part of many painful server migration projects. Users encounter broken shortcuts and drive mappings each time files are relocated. How can you solve this problem permanently? If you use DFS ( What is DFS?), you can virtualize file shares by creating "shares of shares".
How does DFS work? The administrator creates a DFS root that is hosted on a W2K Server in an NT or W2K domain. The root contains the metadata describing the storage namespace. Each DFS link in the namespace points users to a share on a target server. The DFS root simply appears as a share to the users. The DFS links within the root appear as folders when users browse the root.
The DFS client is built into NT, W2K and XP so no desktop touches are required to activate or deploy DFS. If you understand how DNS works, you will definitely understand how DFS works. When the client needs access to a share, it queries the DFS root server. The root server provides a referral to the target server hosting the share. The client caches the referral and then queries the target server for the file in the share.
Subsequent access to the file does not have to query the root server until the cached information times out on the client. If the target share is moved to another server, the DFS information in the root server is updated to point to the new location. When the client queries the root server, it will obtain a referral to the new location. The DFS root server thus acts as a Storage Directory Server, mapping clients to actual share locations, just as DNS maps names to IP addresses.
DFS should be a key strategic component of any organization whose infrastructure includes a large number of Windows file shares.
For a list of DFS technical references, see the following resources:
About the author