Deploying RIS Servers
RIS servers are dependent on your network configuration. The way you deploy and manage your RIS servers on the corporate network determines how your RIS servers perform. By using RIS servers, you can have one operating system image that supports multiple sites, domains, and organizational units, or you can customize each image to meet the needs of the users and computers being served.
You might need multiple RIS servers to support your corporation, or only one RIS server if you are deploying Windows 2000 on a small localized network or network segment. As a general guideline, place a RIS server near the client computers that it services.
The amount of traffic the RIS server produces is similar to that of other servers performing as software distribution points on your network. Generally, the traffic for RIS servers is predictable. RIS-generated traffic is higher when many users are installing their initial operating system image, for example, during a deployment of new operating system images or when a group of new computers is being added to the network. After the operating systems are installed, the daily RIS server traffic will be lower.
Depending on the size of your network, you might need to adjust the distribution and management of client access to RIS servers to streamline access or support multiple operating system requirements. Determine your corporate needs before deploying automated customized versions of Windows 2000 and prior to imaging a standard desktop–configured computer. The number of RIS servers that you need to deploy is determined by the demand for new, upgraded, and customized operating system installations; the speed of your network; and the hardware you use to support your RIS images. Figure 24.2 shows one way to place your RIS servers and optional referral servers in relation to clients for a large organization deployment strategy.
Figure 24.2 Sample RIS Server Layout Within a Large Organization
As illustrated in Figure 24.2, a new remote boot–enabled client requests a remote operating system installation. This request is passed to the RIS referral server, which has the Do not respond to unknown client computers option enabled. Clients that have been prestaged in Active Directory ahead of client servicing can access this RIS server. A prestaged client is a client that already has its computer account object created in Active Directory within a forest. The referral RIS server checks Active Directory to verify whether a computer account object exists for this client. In this example, the client was prestaged by the administrator, therefore it has a corresponding computer account object in Active Directory, and it is assigned to be serviced by RIS Server 3. The RIS referral server passes the request on to RIS server 3, from where the client then begins installing the operating system.
Figure 24.2 shows how one RIS server layout works in a large corporate setting. For this scenario there is close control on which clients can access which RIS servers. When the computer account object is created, the computer account object is assigned to a specific RIS server. Depending on your corporate environment, you can configure your RIS servers so that all RIS servers can respond to all clients. In the network design in Figure 24.2, the only purpose of RIS servers 1, 2, and 3 is to provide images of the operating system. These servers do not respond to initial client service requests. The referral RIS server does not provide image support, however, it does answer client service requests, checks Active Directory for the existence of a prestaged computer account object, and then refers the client to the specified RIS server.
By pre-staging clients to Active Directory and distributing various image files over different distribution points, you can control network traffic and speed up the installation process. Slow connections to your RIS servers can slow down the entire network if they are not designed and distributed appropriately or if the hardware utilized by the RIS server cannot support network demands. If your organization has branch offices, it is best to place a RIS server in each branch location and not attempt to install software over a slow network connection.
RIS does not detect slow links. RIS times out only if it does not receive a DHCP packet from the server.
For more information about optimizing performance of RIS servers, see "Automating Client Installation and Upgrade" in the Deployment Planning Guide.