Hi
Traditionally, a front-end application would be installed on a workstation with an ODBC path to a SQL Server for the data - this is the 'accepted best-practise' way to install applications. Otherwise, you would install the application in Terminal Server/Remote Desktop server etc. etc.
The question I have is, what are the potential pitfalls of installing the application on the server and pointing to the application on the workstation from a UNC path apart from obvious increased network traffic? Is the processing then done on the server or the workstation? This always seemed not an ideal way to run applications, but I cannot find any documentation to support this.
Just telling the client that frontend/backend installation is the correct way just doesn't seem to cut it, when application patch management makes it easier for the client just to install it on the server.
I would rather then install the process under Remote Desktop, but then obviously RDP CAL's become and issue.