ACP Enabled Thin Clients are very flexible and undemanding when added to a network because they are easy to integrate into a wide range of configurations. Through the use of text and illustrations, this paper discusses some sample network configurations using various combinations of Servers and Thin Client hardware.

The simplest network consists of a single terminal server that runs all user applications in self-contained sessions on the terminal server. This is often called a 'Single-Box Network' because there is only one server on the network, and it not only hosts all of the Thin Client sessions but also serves as the platform for ACP ThinManager. Installation and maintenance is simplified by the fact that there is only one PC to install and by the fact that there are no other devices on the network. The server should be configured to include:
The ACP Enabled Thin Clients will boot from the copy of ThinManager that has been installed on the server. Once booted, each Thin Client will then connect back to the same machine (functioning this time as the terminal server), login, and start each individual session.
This scenario is effective for small (1 to 10 client) systems, but larger, more complex networks are improved by dividing the labor among several computers.
Thin Clients can also be configured to run from a terminal server that has been added to a corporate network. The corporate network provides the IP address through DHCP and supplies the network authentication, while the terminal server provides the resources for the Thin Client sessions. This type of configuration begins to offload some of the system 'housekeeping' tasks to other machines on the network, while also giving all of the clients (through the terminal server) access to data throughout the entire plant as shown in the following diagram:

Notice that while some of the licensing has been removed from the Terminal Server (leaving it more available to focus on its primary objective, running user applications) all of the data between the Thin Clients and the terminal server runs on the main network. Unless checked by some other network device, there are therefore many packets roaming around the entire network that are only relevant to the server and the clients. For this reason, many administrators choose to separate the Thin Client network from the other networks.
An additional improvement in a Thin Client system design would be use two network cards in the terminal server, one for the corporate network, and one for the Thin Client network. This lessens the load on the network and allows a separate DHCP Server on the terminal server to be managed independent of the corporate network. The following diagram shows just such an arrangement:

With this design, the administrator has essentially created a subnet that is used only by the Thin Clients. Applications only run on the Terminal Server anyway, so each program simply selects which network it needs to use and only sends packets out of the correct network card.
This Thin Client network design can easily be further enhanced to provide for multiple servers using ACP Failover and redundancy. The hardware to support these standard features of ThinManager can be realized by using two terminal servers, as shown in the figure below. The rainbow-colored boxes represent individual Thin Client sessions.

Each of these Thin Client network configurations has its place, and the only thing that dictates which is the best is the end application. In future issues we will expand these configurations to show how best to adapt them to the installation of various end-user applications.
For more information on ACP Industrial Thin Client computers, please visit our web site at http://www.thinmanager.com
To sign up for the E-mail newsletter go here: ACP newsletter signup
For an archive of past newsletter articles go to: ACP Newsletter Archive