Site home page
(news and notices)

Get alerts when Linktionary is updated

Book updates and addendums

Get info about the Encyclopedia of Networking and Telecommunicatons, 3rd edition (2001)

Download the electronic version of the Encyclopedia of Networking, 2nd edition (1996). It's free!

Contribute to this site

Electronic licensing info



Bootstrapping or Booting

Related Entries    Web Links    New/Updated Information

Search Linktionary (powered by FreeFind)

Note: Many topics at this site are reduced versions of the text in "The Encyclopedia of Networking and Telecommunications." Search results will not be as extensive as a search of the book's CD-ROM.

When a computer is first started, a built-in routine (in ROM) provides it with enough logic to obtain startup programs from a permanent storage device. The built-in ROM routines are small-there is just enough code to direct the system to a disk where it can begin to load the much larger operating system files. But what if the computer is a diskless workstation (a computer with no local storage device)?

In the case of a diskless workstation, boot information is typically obtained from a network server. The diskless workstation has a special ROM inserted on its NIC (network interface card) that directs it to contact a specific computer on the network. This computer then has a program that sends a startup disk image to the diskless workstation so it can boot. All subsequent disk access is performed on network servers.

Network computers need quite a bit of information to get started, including a network address and the location of important services. In the TCP/IP environment, a system needs an IP address, a default router address, a subnet mask, a DNS (Domain Name Service) server address, and some other parameters, depending on the environment. However, creating a startup configuration file for every diskless workstation on a network can be a daunting task, especially if the information changes often, as it might for mobile users. A number of protocols have been developed to dynamically assign IP addresses, as discussed below.

The Internet protocol suite includes a protocol called RARP (Reverse Address Resolution Protocol) that allows a computer to obtain an IP address from a server. When a diskless TCP/IP workstation is booted on a network, it broadcasts a RARP request packet on the local network. This address packet is broadcast on the network for all to receive because the workstation does not know the IP address of the server that can supply it with an address. It includes its own physical network address (the MAC address) in the request so the server will know where to return a reply. The server that receives the request looks in a table,  matches the MAC address with an IP address, and then returns the IP address to the diskless workstation.

Another protocol called BOOTP (BOOTstrap Protocol) provides a way for a server to supply even more configuration information to a workstation at boot time. For more information, see "BOOTP (BOOTstrap Protocol)" and "DHCP (Dynamic Host Configuration Protocol)." DHCP (Dynamic Host Configuration Protocol) can provide automatic and dynamic IP address allocation for some or all of the workstations on a network.

Copyright (c) 2001 Tom Sheldon and Big Sur Multimedia.
All rights reserved under Pan American and International copyright conventions.