4.3BSD/usr/doc/smm/11.named/build.me

.\" Copyright (c) 1986 Regents of the University of California.
.\" All rights reserved.  The Berkeley software License Agreement
.\" specifies the terms and conditions for redistribution.
.\"
.\"	@(#)build.me	6.1 (Berkeley) 5/7/86
.\"
.sh 1 "Building A System with a Name Server"
.pp
BIND is comprised of two parts.  One is the user interface called the 
\fIresolver\fP
which consists of a group of routines that reside in the C library 
\fI/lib/libc.a\fP.
Second is the actual server called \fInamed\fP.
This is a daemon that runs in the background and services queries on a 
given network port. The standard port for UDP and TCP is specified in 
\fI/\|etc/\|services\fP.
.sh 2 "Resolver Routines in libc"
.pp
When building your 4.3BSD system you may either
build the C library to use the name server resolver routines 
or use the host table lookup routines to do host name and address resolution.
The default resolver for 4.3BSD uses the name server.
.pp
Building the C library to use the name server changes the way
\fIgethostbyname\fP\|(3N), \fIgethostbyaddr\fP\|(3N), and \fIsethostent\fP\|(3N)
do their functions.
The name server renders \fIgethostent\fP\|(3N) obsolete,
since it has no concept of a next line in the database.
These library calls are built with the resolver routines needed
to query the name server.
.pp
The \fIresolver\fP is comprised of a few routines that build query
packets and exchange them with the name server.
.pp
Before building the C library, set the variable \fIHOSTLOOKUP\fP
equal to \fInamed\fP in \fI/\|usr/\|src/\|lib/\|libc/\|Makefile\fP. 
You then make and install the C library and compiler and then
compile the rest of the 4.3BSD system.  For more information
see section 6.6 of ``Installing and Operating 4.3BSD on the VAX\(dd''.
.(f
\(ddVAX is a Trademark of Digital Equipment Corporation
.)f

.sh 2 "The Name Service"
.pp
The basic function of the name server is to provide information about network
objects by answering queries.  The specifications for this name server
are defined in RFC882, RFC883, RFC973 and RFC974.
These documents can be found in \fI/usr/src/etc/named/doc\fP in 4.3BSD 
or \fIftp\fPed from sri-nic.arpa. It is also recommeded that 
you read the related
manual pages,  \fInamed\fP\|(8),
\fIresolver\fP\|(3),
and \fIresolver\fP\|(5).
.pp
The advantage of using a name server over the host table lookup for
host name resolution is to avoid the need 
for a single centralized clearinghouse for all names.
The authority for this information can be delegated 
to the different organizations on the network responsible for it.
.pp
The host table lookup routines require that the master file
for the entire network be maintained at a central location by a few people.
This works fine for small networks where there are only a few machines and the
different organizations responsible for them cooperate.
But this does not work well for large networks where machines
cross organizational boundaries.
.pp
With the name server, the network can be broken into a hierarchy of domains. 
The name space is organized as a tree according to organizational or
administrative boundaries. 
Each node, called a \fIdomain\fP, is given a label, and the name of the
domain is the concatenation of all the labels of the domains from
the root to the current domain, listed from right to left separated by dots.
A label need only be unique within its domain.
The whole space is partitioned into several areas called \fIzones\fP,
each starting at a domain and extending down to the leaf domains or to
domains where other zones start.  
Zones usually represent administrative boundaries.
An example of a host address for a host at the University of California,
Berkeley would look as follows:
.(b
\fImonet\fP\|\fB.\fP\|\fIBerkeley\fP\|\fB.\fP\|\fIEDU\fP
.)b
The top level domain for educational organizations is EDU;
Berkeley is a subdomain of EDU and monet is the name of the host.