Terminal Services Part 7: Positioning TS Servers
The positioning of your Terminal Servers is also another important consideration when planning your rollout. Client/server applications tend to assume that there is a high level of network connectivity and bandwidth between their “client” and “server” halves. When that is not the case – like what happens when users attempt to access applications remotely or when latent network connections are used to remote offices – client/server applications suffer in terms of performance.
To best explain this behavior, consider the Outlook client. When connected to an Exchange server, this client has a high expectation of good network performance between the client and server. But there are times when that Outlook client cannot be in close network proximity to its server, such as when the user is in a remote office or is working over a VPN line. In either of these cases, Outlook is likely to experience low performance because the network cannot provide for the load required by the client.
One very effective use of Terminal Services is in just these circumstances. By installing a Terminal Server in close network proximity to the “server” half of a client/server application, the Terminal Server can handle the bandwidth-intensive client-to-server communication. At the same time, it will serve up the user interface to the user through the not-intensive RDP protocol.Â
Always keep in mind the low-bandwidth and/or high-latency benefits of the RDP protocol and how it impacts where you may position your Terminal Servers. Be conscious about where you position your Terminal Servers in relation to those servers your applications work with as well as the position of users who run your applications.
The RDP protocol used by Terminal Services is “thinner” than most protocols used by client/server applications today. As such, it can often be an excellent solution for hosting applications for remote offices and users when the performance of local clients just won’t work for your users.




