Want to map a folder as a drive letter

[Kenazo]

We run a Windows 2008 R2 remote desktop server and I want our users to have access to a folder on the server's hard drive as a drive letter.

Are there any benefits to doing this using the SUBST command versus just setting it up as a network share and setting it up as a mapped drive?

Is there an even better way to do this?

 

[JackMDS]

If the Network resource is unavailable (server is down, etc.)

SUBST will continually retry.

NET USE which will try to connect once and if it fials it is out of the system.

 

[imagoon]

NET USE driveletter: \\servername\share\folder /persistent:yes

works best when the login script is this:

NET USE driveletter: /delete
NET USE driveletter: \\servername\share\folder /persistent:yes

this deletes the share before recreating it during the login process. Not required but it tends to eliminate strangeness.

SUBST is old school and I am surprised it still works really.

Something to think about... when someone is logged in to the desktop even via RDP there is some security concerns that might make it wise to push the data to a second server that is easier to secure. You can secure the server desktop pretty well but it just takes one opening or accident and you have a program rampaging the server.

I also highly recommend looking at DFS and DFS-R while using 2k8 R2.

 

[imagoon]

If the Network resource is unavailable (server is down, etc.)

SUBST will continually retry.

NET USE which will try to connect once and if it fials it is out of the system.

Not exactly. Persistent connections stay until deleted. This is how I keep the drive letters there for the VPN users. The reason I do the order I do in the log in script is this: When a user shuts down and goes home and powers backup the persistent connections are still there, they can VPN in and the drive letters automagically work because the login scripts will not run remotely. However in the office in the morning the systems get refreshed which is good if I need to make a change.

This happens far less now however because I use Distributed Files System (DFS) to handle data share routing. I can move shares around on servers as needed and "i: \software" is always the same as far as the clients care. It is really worth while looking in to and using.

 

[Kenazo]

If the Network resource is unavailable (server is down, etc.)

SUBST will continually retry.

NET USE which will try to connect once and if it fials it is out of the system.

The network resource being unavailable will never be a problem since the hard drive is physically located on an array in the RDS box. It has a RAID 1 array for programs & OS and a RAID 6 array for DATA. There are a number of items on the RAID6 array that I'd prefer users aren't accidentally accessing which is why I was thinking of just mapping our \data\ directory on that drive as its own drive letter.

We are currently using NET USE to do this, but this means that all of our programs think this is a network share, not a local disk. I'm trying to get the computer to think that this data folder is on a local disk, but still have it show as its own drive letter.

 

[Kenazo]

Something to think about... when someone is logged in to the desktop even via RDP there is some security concerns that might make it wise to push the data to a second server that is easier to secure. You can secure the server desktop pretty well but it just takes one opening or accident and you have a program rampaging the server.

I also highly recommend looking at DFS and DFS-R while using 2k8 R2.

We previously had our data on a Linux file server but found that the gigabit connection was just too slow for our needs. Some of our programs (Caseware in particular) felt very laggy b/c they were continually reading/writing to hundreds of small files on the data server. Now we have everything local and it's lightning fast. 🙂

 

[imagoon]

We are currently using NET USE to do this, but this means that all of our programs think this is a network share, not a local disk. I'm trying to get the computer to think that this data folder is on a local disk, but still have it show as its own drive letter.

Why does this matter?

 

[RebateMonger]

I'm trying to get the computer to think that this data folder is on a local disk, but still have it show as its own drive letter.

I'm not sure it's that easy to do. Otherwise, folks would use it in Hyper-V for all kinds of stuff that can't be done with a networked drive. iSCSI will do it, though.

 

[Kenazo]

Why does this matter?

A few of our programs give us warnings when they access data across a network, but by using SUBST (which makes the drive appear to be a local one, which in our case it is) the program doesn't give these warnings. It works fine either way, just trying to avoid users getting warning messages.

 

[corkyg]

Would not the simple solution be to create a partition and give it the folder name? It would then automagically be a lettered drive.