RIS Server PXE Environment

RIS uses DHCP as part of what is defined in the PXE architecture to initiate the process of remotely installing an operating system on a client. In other words, because PXE uses DHCP, so does RIS.

When a new DHCP PXE-based remote boot client is turned on for the first time, the client requests an IP address and the IP address of an active RIS server through the DHCP protocol and the PXE extensions to the DHCP protocol. As part of the initial request, as a DHCP option, the client sends out its GUID, which is used to identify the client in Active Directory. The client receives an IP address from the DHCP server and the IP address of the RIS server that services the client. In the RIS server's response, the client is given the name of a boot image that it must request when contacting the RIS server for initial service.

When the client makes its initial request for service, TFTP is used to download the boot image file to the client. In the case of RIS, this file is Startrom.com. Startrom.com prompts user to press the F12 key. If the user presses the F12 key, Startrom.com uses TFTP to download OSChooser, and presents the user with the Client Installation Wizard.

The process of initial communication between PXE clients and RIS servers can differ depending on how RIS is deployed in relation to DHCP services.

DHCP and RIS on Separate Servers

If you have DHCP and RIS on separate servers, the initial interaction between PXE clients and RIS/DHCP servers proceeds as follows:

1. DHCP discover from client (asking for IP address and PXE boot server).
2. DHCP offer from DHCP server (offers IP address and other network configuration settings).
3. DHCP offer from RIS server (offers PXE boot server).
4. DHCP request from client to DHCP server (requesting IP address).
5. DHCP acknowledge message from DHCP server (you can have this IP address).
6. DHCP request from client to RIS server (requesting the boot server).
7. DHCP acknowledge message from RIS server (this acknowledgment contains the address to the RIS server and the first file that the client needs to send a TFTP request to start the boot process).

Note

If you configure the RIS server to respond only to known clients — that is, clients prestaged in Active Directory or previously installed computers — and the computer object is not located in Active Directory, the RIS server fails to respond to the client's DHCP request. If the RIS server is not on the same server as the DHCP server, and the server does not respond because the client is unknown, then the DHCP offer from the RIS server (in step 3 in the previous process) is not sent and therefore step 6 and step 7 do not occur.

DHCP and RIS on the Same Server

If the RIS server and the DHCP server are on the same computer, the conversation is as follows:

1. DHCP discover from client (asking for IP address and PXE boot server).
2. DHCP offer from DHCP/RIS server (offers IP address and PXE boot server).
3. DHCP request from client to DHCP server (requesting IP address, network configuration settings, and PXE boot server).
4. DHCP acknowledge from DHCP server (contains IP address and the RIS server IP and the first file to download).

Note

If you configure the RIS server to respond only to known clients, and the computer object is not located in Active Directory, the RIS server fails to respond to the client's DHCP request. If the RIS server and DHCP server are on the same computer, the DHCP offer from the DHCP/RIS server (in step 2 in the previous process) only contains IP information and no information about any available servers to support the client's network boot process.

© 1985-2000 Microsoft Corporation. All rights reserved.