You probably already know that Windows 2000 contains a wide variety of tools that you can use to see what's going on inside your servers. For example, you can use the Event Viewer Console to see if any warnings or errors are present on a server. However, even with all these tools, sometimes there's no substitute for being able to actually see the server's screen in real time. Windows 2000's terminal services will allow you to do just that from a wide variety of clients.
In this series of articles, I'll explain how to manage your servers through a terminal-enabled client. In Part 1, I'll discuss the basics of the terminal services. In future articles, I'll discuss enabling the terminal services on your servers and how to prepare a client to access the terminal services.
It's often difficult for administrators to grasp the concept of the terminal services, because the terminal services perform more like something from the mainframe world than from the PC world. Dozens of third-party products offer remote-control capabilitiesbut most, if not all, of these third-party remote control products tie up the actual server console during use. If the remote user moves his mouse, anyone who's sitting at the server console will also see the server's mouse pointer move. Needless to say, using traditional remote control packages on a server can compromise your network's security if you aren't extremely careful about how the remote control capabilities are implemented.
The terminal services work differently than traditional remote control software. When a terminal-enabled client attaches to the server, it doesn't see the actual server screen, but rather a reproduction of it. Therefore, if a terminal-enabled client is at work, the server screen appears as though no one is using it. Of course, there are ways to check to see if someone is using a terminal session, but the idea is that a terminal-enabled client doesn't take full control of the server.
To understand exactly how the terminal services work, think of a mainframe environment in which a central computer sends screen updates to dumb terminals. The mainframe itself performs every bit of the processing. Any time a user enters information, the information is sent to the mainframe and processed, and a screen update is passed back to the dumb terminal to acknowledge the processing. This concept is very similar to the way that the terminal services work.
Another similarity is that in a mainframe environment, everyone who's attached to the mainframe has her own individual session. This makes it possible for everyone to work independently. Just because the mainframe is sending everyone screen updates doesn't mean all the screens are the same. Instead, the mainframe maintains an independent session for each attached user, which allows users to work independently of each other.
What Are the Terminal Services?
You Decide How to Use the Services
Right about now, you might be getting a little concerned. PC users tend to have a negative image of mainframes, because mainframes are ponderous, cost millions of dollars, and are usually reserved for use as storage systems with millions of records. And let's face itno one likes to think of their high-end Pentium III as a glorified dumb terminal.
Fortunately, if you decide to use the terminal services, how you use them is up to you. You're not limited to an all-or-nothing situation. As I'll discuss in detail in Part 2, when you initially set up the terminal services, you're asked whether the server will function as a full-blown terminal server, or if you just plan to use the terminal services for administrative purposes.