ULTRIX-32 Guide to the BIND Service

This document is the "ULTRIX-32 Guide to the BIND Service," version 3.0, published by Digital Equipment Corporation in 1987-1988.

Purpose: The guide introduces the Berkeley Internet Name Domain (BIND) service, explaining how to install, configure, manage, and troubleshoot it on the ULTRIX operating system. It provides comprehensive information for network and system administrators to implement decentralized host name and address management.

Audience: It is intended for system managers, network managers, or MicroVAX users running ULTRIX, who are familiar with ULTRIX commands, system configuration, naming conventions, and local network topology.

Key Content:

1. Introduction to BIND:

   • What BIND is: A host name and address lookup service for the Internet, acting as an alternative or supplement to /etc/hosts or Yellow Pages (YP). It comprises a "resolver" (client-side software) and a "server" (the named daemon).
   • Advantages: Enables decentralized management of host information, ideal for large, organizationally diverse networks.
   • Concepts: Explains the hierarchical structure of domains and zones, including top-level domains like .com, .edu, .gov, etc.
   • BIND Server Types: Details various server roles:
     • Root Servers: Ultimate authorities for top-level domains.
     • Master Servers (Primary/Secondary): Maintain BIND databases for their zones, with secondary servers providing redundancy and data synchronization.
     • Caching Servers: Cache information to speed up lookups but hold no authoritative data.
     • Forwarding Servers: Resolve recursive queries for slave servers, providing Internet access without direct interaction from all hosts.
     • Slave Servers: Forward queries to configured forwarders if they cannot resolve them locally.
   • BIND Clients: Systems that use BIND for name resolution via the /etc/resolv.conf file, without running the named daemon.
   • Query Resolution: Outlines the step-by-step process BIND servers follow to resolve host name and IP address queries.

2. Setting Up BIND:

   • Provides instructions for automatically and manually setting up BIND clients and servers using the bindsetup command or by editing configuration files.
   • Explains the purpose and format of key BIND-related files, including named.boot (server type and data file locations), and data files like named.ca (root server hints), named.local (loopback interface), named.hosts (host/address info), and named.rev (reverse address mapping).
   • Details various BIND file entry types (Resource Records - RR) such as SOA (Start of Authority), NS (Name Server), A (Address), HINFO (Host Information), WKS (Well Known Services), CNAME (Canonical Name), PTR (Pointer), MB (Mailbox), MR (Mail Rename), MINFO (Mailbox Information), and MX (Mail Exchanger).

3. Managing and Using BIND:

   • Covers responsibilities of Domain Administrators (DA) and Technical/Zone Contacts.
   • Explains naming domains and hosts and the process of registering a domain with public networks like DARPA Internet (ARPANET), CSNET, and BITNET.
   • Describes how to update BIND data files and the importance of incrementing the serial number in SOA records.
   • Introduces utility commands like nslookup and nsquery for obtaining host name and IP address information, and the NIC whois service for domain contact details.

4. Troubleshooting BIND:

   • Offers guidelines for diagnosing and resolving common BIND issues, primarily by reviewing configuration and debug files (/var/spool/mqueue/syslog, /var/tmp/named_dump.db, /var/tmp/named.run, /tmp/named.stat).
   • Explains how to obtain the named daemon's process ID and send signals (e.g., SIGHUP for reloading data, SIGINT for dumping the database, USR1 for debugging, KILL for termination).

Appendices: Include sample BIND files, a BIND registration questionnaire, a list of related papers and RFCs, and detailed nslookup command examples for practical use.