[Top] [Prev] [Next] [Contents] [Index]

Using RARP for Address Discovery

You can use RARP instead of BOOTP/DHCP to determine addresses; however, RARP returns only the IP address of the terminal and the boot host. You must set other addresses needed by the terminal, such as the gateway and subnet mask, in NVRAM. For information about configuring addresses in NVRAM, see "Storing Addresses in NVRAM".

You cannot use RARP to specify the X server file to load; instead, the terminal uses the default download sequence to locate an X server. For information about the download sequence, see Chapter 4.

RARP is implemented through a daemon program, rarpd(8) that runs on the boot host and a database file called /etc/ethers.

Complete the following tasks if you are using RARP for address discovery:

  1. To verify that RARP is available, check the relevant host startup file. For example, on SunOS systems, the startup file to check is /etc/rc.local. The entry for RARP is similar to:

    rarpd -a

  2. If the entry is disabled by a comment symbol (#) at the beginning of the line, remove the comment symbol. Then, start the daemon manually by typing the startup command. For example:

    # rarpd
    # rarpd -a 

  3. If you are not running NIS (Network Information Service), add each terminal's Ethernet address and hostname to the /etc/ethers file. For example:

    00:00:A7:00:00:AE ncd1

    The Ethernet address is entered into NVRAM at the factory. You can display the address through the Console (Statistics -> Show Version). The Ethernet address also appears on a label on the bottom of the terminal base.

  4. If you are running NIS, add the terminal to the ethers map and update the map. For example:

    # cd /var/yp
    # make ethers

[Top] [Prev] [Next] [Contents] [Index]

Send comments, suggestions, or questions about this document to the NCD Technical Publications Department by Internet e-mail. Write to us at techpubs@ncd.com.
Copyright © 1997, NCD Inc. All rights reserved.