297

C H A P T E R 5

Introduction to DNS

Domain Name System (DNS) enables you to use hierarchical, friendly names to easily
locate computers and other resources on an IP network. The following sections
describe the basic DNS concepts, including features explained in newer Requests for
Comments (RFCs), such as dynamic update, from the Internet Engineering Task Force
(IETF). The Microsoft® Windows® 2000–specific implementation of DNS is not
covered within this chapter, except where indicated.
For information about the Windows 2000 implementation of DNS, see
“Windows 2000 DNS” in this book.
DNS is a distributed database that contains mappings of DNS domain names to data.
It is also a protocol for Transmission Control Protocol/Internet Protocol (TCP/IP)
networks, defined by the Requests for Comments (RFCs) that pertain to DNS. DNS
defines the following:
• Mechanism for querying and updating the database.
• Mechanism for replicating the information in the database among servers.
• Schema for the database.

In This Chapter
Introduction to the Domain Name System 299
DNS Servers 306
Name Resolution 309
Resource Records and Zones 312
Zone Transfer 328
Dynamic Update 331
DNS Standards 332

Related Information in the Resource Kit

• For more information about TCP/IP protocols, see “Introduction to TCP/IP” in
this book.
• For information about the Windows 2000 implementation of DNS, see
“Windows 2000 DNS” in this book.
298 Part 2 Address Allocation and Name Resolution

Introduction to the Domain Name System

Although TCP/IP uses IP addresses to locate and connect to hosts (computers and
other TCP/IP network devices), users typically prefer to use friendly names. For
example, users prefer the friendly name ftp.reskit.com, instead of its IP address,
172.16.23.55. The Domain Name System (DNS), defined in RFCs 1034 and 1035, is
used on the Internet to provide a standard naming convention for locating IP-based
computers.
On the Internet, before the implementation of DNS, the use of names to locate
resources on TCP/IP networks was supported by a file called Hosts. Network
administrators entered names and IP addresses into Hosts, and computers used the file
for name resolution.
Both the Hosts file and DNS use a namespace. A namespace is a grouping in which
names can be used to symbolically represent another type of information, such as an
IP address, and in which specific rules are established that determine how names can
be created and used. Some namespaces, such as DNS, are hierarchically structured
and provide rules that allow for the namespace to be divided into subsets of names for
distributing and delegating parts of the namespace. Other namespaces, such as the
Hosts namespace cannot be divided and must be distributed in their entirety. Because
of this, using the Hosts file posed a problem for network administrators. As the
number of computers and users on the Internet grew, the task of updating and
distributing the Hosts file became unmanageable.
DNS replaces the Hosts file with a distributed database that implements a hierarchical
naming system. This naming system allows for growth on the Internet and the creation
of names that are unique throughout the Internet and private TCP/IP-based intranets.

Domain Namespace
The naming system on which DNS is based is a hierarchical and logical tree structure
called the domain namespace. Organizations can also create private networks that are
not visible on the Internet, using their own domain namespaces. Figure 5.1 shows part
of the Internet domain namespace, from the root domain and top-level Internet DNS
domains, to the fictional DNS domain named reskit.com that contains a host
(computer) named Mfgserver.
 Chapter 5 Introduction to DNS 299

Managed by
Internet authority

“ ”
(Root)

com Other
Top-level
edu org Domain

reskit Top-level Internet

Domains

Mfgserver

Reskit
Domain

Figure 5.1 Domain Name System

Each node in the DNS tree represents a DNS name. Some examples of DNS names
are DNS domains, computers, and services. A DNS domain is a branch under the
node. For example, in Figure 5.1, reskit.com is a DNS domain. DNS domains can
contain both hosts (computers or services) and other domains (referred to as
subdomains). Each organization is assigned authority for a portion of the domain
namespace and is responsible for administering, subdividing, and naming the DNS
domains and computers within that portion of the namespace.
Subdividing is an important concept in DNS. Creating subdivisions of the domain
namespace and private TCP/IP network DNS domains supports new growth on the
Internet and the ability to continually expand name and administrative groupings.
Subdivisions are generally based on departmental or geographic divisions.
For example, the reskit.com DNS domain might include sites in North America and
Europe. A DNS administrator of the DNS domain reskit.com can subdivide the
domain to create two subdomains that reflect these groupings: noam.reskit.com. and
eu.reskit.com. Figure 5.2 shows an example of these subdomains.
300 Part 2 Address Allocation and Name Resolution

“ ”
(Root)

com

reskit

noam eu

Figure 5.2 Subdomains

Domain Name
Computers and DNS domains are named based on their position in the domain tree.
For example, because reskit is a subdomain of the .com domain, the domain name for
reskit is reskit.com.
Every node in the DNS domain tree can be identified by a fully qualified domain
name (FQDN). The FQDN is a DNS domain name that has been stated
unambiguously so as to indicate with absolute certainty its location relative to the root
of the DNS domain tree. This contrasts with a relative name, which is a name relative
to some DNS domain other than the root.
For example, the FQDN for the server in the reskit.com DNS domain is constructed as
Mfgserver.reskit.com., which is the concatenation of the host name (Mfgserver) with
the primary DNS suffix (reskit.com), and the trailing dot (.). The trailing dot is a
standard separator between the top-level domain label and the empty string label
corresponding to the root.
Note In general, FQDNs have naming restrictions that allow only the use of
characters a-z, A-Z, 0-9, and the dash or minus sign (-). The use of the period
(.) is allowed only between domain name labels (for example, “reskit.com”) or at the
end of a FQDN. Domain names are not case-sensitive.
You can configure the Windows 2000 DNS server to enforce some or all RFC
character restrictions or to ignore all character restrictions. For more information, see
“Windows 2000 DNS” in this book.

Internet Domain Namespace

The root (the top-most level) of the Internet domain namespace is managed by an
Internet name registration authority, which delegates administrative responsibility for
portions of the domain namespace to organizations that connect to the Internet.
Beneath the root DNS domain lie the top-level domains, also managed by the Internet
name registration authority. There are three types of top-level domains:
• Organizational domains. These are named by using a 3-character code that
indicates the primary function or activity of the organizations contained within
 Chapter 5 Introduction to DNS 301

the DNS domain. Organizational domains are generally only for organizations
within the United States, and most organizations located in the United States
are contained within one of these organizational domains.
• Geographical domains. These are named by using the 2-character
country/region codes established by the International Standards Organization
(ISO) 3166.
• Reverse domains. This is a special domain, named in-addr.arpa, that is used for
IP address-to-name mappings (referred to as reverse lookup). For more
information, see “Name Resolution” later in this chapter. There is also a
special domain, named IP6.INT, used for IP version 6 reverse lookups. For
information, see RFC 1886.
The most commonly used top-level DNS name components for organizations in the
United States are described in the Table 5.1.
Table 5.1 Top-Level Name Component of the DNS Hierarchy
Top-Level
Name Example DNS
Component Description Domain Name
.com An Internet name authority delegates portions of the microsoft.com
domain namespace under this level to commercial
organizations, such as the Microsoft Corporation.
.edu An Internet name authority delegates portions of this mit.edu
domain namespace to educational organizations, such
as the Massachusetts Institute of Technology (MIT).
.gov An Internet name authority delegates portions of this whitehouse.gov
domain namespace to governmental organizations,
such as the White House in Washington, D.C.
.int An Internet name authority delegates portions of this nato.int
domain namespace to international organizations, such
as the North Atlantic Treaty Organization (NATO).
.mil An Internet name authority delegates portions of this ddn.mil
domain namespace to military operations, such as the
Defense Date Network (DDN).
(continued)
Table 5.1 Top-Level Name Component of the DNS Hierarchy (continued)
Top-Level
Name Example DNS
Component Description Domain Name
.net An Internet name authority delegates portions of this nsf.net
domain namespace to networking organizations, such
as the National Science Foundation (NSF).
.org An Internet name authority delegates portions of this cnidr.org
domain namespace to noncommercial organizations,
such as the Center for Networked Information
302 Part 2 Address Allocation and Name Resolution

Discovery and Retrieval (CNIDR).

In addition to the top-level domains listed above, individual countries have their own
top-level domains. For example, .ca is the top-level domain for Canada.
Beneath the top-level domains, an Internet name authority delegates domains to
organizations that connect to the Internet. The organizations to which an Internet
name authority delegates a portion of the domain namespace are then responsible for
naming the computers and network devices within their assigned domain and its
subdivisions. These organizations use DNS servers to manage the name-to-IP address
and IP address-to-name mappings for host devices contained within their portion of
the namespace.

Basic DNS Concepts

This section provides brief definitions of additional DNS concepts, which are
described in more detail in the following sections of this chapter.
DNS servers. Computers that run DNS server programs containing DNS database
information about the DNS domain tree structure. DNS servers also attempt to resolve
client queries. When queried, DNS servers can provide the requested information,
provide a pointer to another server that can help resolve the query, or respond that it
does not have the information or that the information does not exist.
DNS resolvers. Programs that use DNS queries to query for information from servers.
Resolvers can communicate with either remote DNS servers or the DNS server
program running on the local computer. Resolvers are usually built into utility
programs or are accessible through library functions. A resolver can run on any
computer, including a DNS server.
Resource records. Sets of information in the DNS database that can be used to process
client queries. Each DNS server contains the resource records it needs to answer
queries for the portion of the DNS namespace for which it is authoritative. (A DNS
server is authoritative for a contiguous portion of the DNS namespace if it contains
information about that portion of the namespace.)
Zones. Contiguous portions of the DNS namespace for which the server is
authoritative. A server can be authoritative for one or more zones.
Zone files. Files that contain resource records for the zones for which the server is
authoritative. In most DNS implementations, zones are implemented as text files.

Zones
A zone is a contiguous portion of the DNS namespace. It contains a series of records
stored on a DNS server. Each zone is anchored at a specific domain node. However,
zones are not domains. A DNS domain is a branch of the namespace, whereas a zone
is a portion of the DNS namespace generally stored in a file, and can contain multiple
domains. A domain can be subdivided into several partitions, and each partition, or
zone, can be controlled by a separate DNS server. Using the zone, the DNS server
answers queries about hosts in its zone, and is authoritative for that zone. Zones can
be primary or secondary. A primary zone is the copy of the zone to which the updates
 Chapter 5 Introduction to DNS 303

are made, whereas a secondary zone is a copy of the zone that is replicated from a
master server.
Zones can be stored in different ways. For example, they can be stored as zone files.
On Windows 2000 servers, they can also be stored in the Active Directory™ directory
service. Some secondary servers store them in memory and perform a zone transfer
whenever they are restarted.
Figure 5.3 shows an example of a DNS domain that contains two primary zones. In
this example, the domain reskit.com contains two subdomains: noam.reskit.com. and
eu.reskit.com. Authority for the noam.reskit.com. subdomain has been delegated to
the server noamdc1.noam.reskit.com. Thus, as Figure 5.3 shows, one server,
noamdc1.noam.reskit.com, hosts the noam.reskit.com zone, and a second server,
reskitdc1.reskit.com, hosts the reskit.com zone that includes the eu.reskit.com
subdomain.
“ ”

com

reskit

reskitdc1
eu
noam

noamdc1

reskit.com
zone
noam.reskit.com
zone

Figure 5.3 Domains and Zones

Rather than delegating the noam.reskit.com zone to noamdc1.noam.reskit.com, the

administrator can also configure reskitdc1 to host the zone for noam.reskit.com.
Also, you cannot configure two different servers to manage the same primary zones;
only one server can manage the primary zone for each DNS domain. There is one
exception: multiple computers can manage Windows 2000 Active Directory–
integrated zones. For more information, see “Windows 2000 DNS” in this book.
You can configure a single DNS server to manage one zone or multiple zones,
depending on your needs. You can create multiple zones to distribute administrative
tasks to different groups and to provide efficient data distribution. You can also store
the same zone on multiple servers to provide load balancing and fault tolerance.
304 Part 2 Address Allocation and Name Resolution

For information about what zones contain, see “Resource Records and Zones” later in
this chapter.

DNS Servers
DNS servers store information about no zones, one zone, or multiple zones. When a
DNS server receives a DNS query, it attempts to locate the requested information by
retrieving data from its local zones. If this fails because the server is not authoritative
for the DNS domain requested and thus does not have the data for the requested
domain, the server can check its cache, communicate with other DNS servers to
resolve the request, or refer the client to another DNS server that might know the
answer.
DNS servers can host primary and secondary zones. You can configure servers to host
as many different primary or secondary zones as is practical, which means that a
server might host the primary copy of one zone and the secondary copy of another
zone, or it might host only the primary or only the secondary copy for a zone. For
each zone, the server that hosts the primary zones is considered the primary server for
that zone, and the server that hosts the secondary zones is considered the secondary
server for that zone.
Primary zones are locally updated. When a change is made to the zone data, such as
delegating a portion of the zone to another DNS server or adding resource records in
the zone, these changes must be made on the primary DNS server for that zone, so
that the new information can be entered in the local zone.
In contrast, secondary zones are replicated from another server. When a zone is
defined on a secondary server for that zone, the zone is configured with the IP address
of the server from which the zone is to be replicated. The server from which the zone
file replicates can either be a primary or secondary server for the zone, and is
sometimes called a master server for the secondary zone.
When a secondary server for the zone starts up, it contacts the master server for the
zone and initiates a zone transfer. The secondary server for the zone also periodically
contacts the master server for the zone to see whether the zone data has changed. If so,
it can initiate a transfer of the zones, referred to as a zone transfer. For more
information about zone transfers, see “Zone Transfer” later in this chapter.
You must have a primary server for each zone. Additionally, you should have at least
one secondary server for each zone. Otherwise, if the primary server for the zone goes
down, no one will be able to resolve the names in that zone.
Secondary servers provide the following benefits:
Fault tolerance When a secondary server is configured for a zone, clients can still
resolve names for that zone even if the primary server for the zone
goes down. Generally, plan to install the primary and secondary servers for the zone
on different subnets. Therefore, if connectivity to one subnet is lost, DNS clients can
still direct queries to the name server on the other subnet.
Reduction of traffic on wide area links You can add a secondary server for
the zone in a remote location that has
a large number of clients, and then configure the client to try those servers first. This
can prevent clients from communicating across slow links for DNS queries.
Reduction of load on the primary server for the zone The secondary
 Chapter 5 Introduction to DNS 305

server can answer queries for the zone, reducing the number of queries the primary
server for the zone must answer.
The following sections describe servers that act as caching-only servers, forwarders,
and slaves.

Caching-Only Servers
All DNS servers perform caching; whenever they receive information from other
servers, they store the information for a certain amount of time. This speeds the
performance of DNS resolution, reduces DNS-related query traffic, and improves
reliability. For more information, see “Caching and Time to Live” later in this chapter.
Certain DNS servers, known as caching-only servers, simply perform queries, cache
the answers, and return the results. They are not authoritative for any DNS domains
and do not host any zones. They only store data that they have cached while resolving
queries.
The benefit provided by caching-only servers is that they do not generate zone
transfer network traffic because they do not contain any zones. However, there is one
disadvantage: when the server is initially started, it has no cached information and
must build up this information over time as it services requests.

Forwarders and Slaves

When a DNS server receives a query, it attempts to locate the requested information
within its local zones and from the cache. If it cannot locate the requested information
and is not authoritative for the requested information, it must communicate with other
servers to resolve the request. However, in some cases network administrators might
not want the server to communicate directly with other servers. For example, if your
organization were connected to the Internet by means of a slow wide area link, you
might not want every DNS server in your organization to connect directly to DNS
servers on the Internet.
To solve this problem, DNS allows for the use of forwarders. Forwarders are DNS
servers that are designated to provide forwarding of off-site queries for other DNS
servers. For example, you could designate one DNS server as a forwarder for names
of computers on the Internet, and then configure your other servers to use that
forwarder to resolve names for which they are not authoritative.
You do not need to perform any special configuration on the computer designated as a
forwarder. You must configure the DNS server that needs to forward queries by
providing the IP address of the forwarders.
A server can use a forwarder in a nonexclusive or exclusive mode. In a nonexclusive
mode, when a server receives a DNS query for which it is not authoritative and cannot
resolve through its own zones or cache, it passes the query to one of the designated
forwarders. The forwarder then carries out whatever communication is necessary to
resolve the query and returns the results to the requesting server, which returns the
results to the original requester. If the forwarder cannot resolve the query, the server
that received the original query attempts to resolve the query on its own.
In an exclusive mode, servers rely completely on the name-resolving ability of the
forwarders. Servers using forwarders in an exclusive mode are known as slaves.
306 Part 2 Address Allocation and Name Resolution

When a slave receives a DNS query that it cannot resolve through its own zones, it
passes the query to one of the designated forwarders. The forwarder then carries out
whatever communication is necessary to resolve the query and returns the results to
the slave, which returns the results to the original requester. If the forwarder cannot
resolve the request, the slave returns a query failure to the original requestor. Slaves
make no attempt to resolve the query on their own if the forwarder cannot satisfy the
request.

Load Sharing
DNS servers use a mechanism called round-robin or load sharing, explained in RFC
1794, to share and distribute loads for network resources. Round-robin rotates the
order of resource record data returned in a query answer in which multiple RRs exist
of the same RR type for a queried DNS domain name.
For example, suppose you have three World Wide Web servers with the same domain
name, WWWServer, that all display the Web page for www.reskit.com, and you want
to share the load between them. On the name server, you would create the following
resource records:
www.reskit.com. IN A 172.16.64.11
www.reskit.com. IN A 172.17.64.22
www.reskit.com. IN A 172.18.64.33

A name server configured to perform round-robin rotates the order of the A resource
records when answering client requests. In this example, the name server would reply
to the first client request by ordering the addresses as 172.16.64.11, 172.17.64.22, and
172.18.64.33. It would reply to the second client response by ordering the addresses
as 172.17.64.22, 172.18.64.33, and 172.16.64.11. The rotation process continues until
data from all of the same type of resource records associated with a name have been
rotated to the top of the list returned in answering client queries. The client is required
to try the first IP address listed.
By default, a Windows 2000 DNS server uses a different method to order the records
returned to a client. It attempts to find the resource record containing the IP address
closest to the client, then returns this resource record first in the list of records.
However, you can modify the default so it performs traditional round-robin. For
information, see “Windows 2000 DNS” in this book. Versions of BIND 4.9.3 and
later perform this type of load sharing. Earlier versions of BIND performed a different
type of load sharing; for more information, see RFC 1794.

Name Resolution
DNS clients use libraries called resolvers that perform DNS queries to servers on
behalf of the client. Keep in mind throughout this discussion that a DNS server can
also be a client to another server.
Note Computers running under Microsoft® Windows NT® Workstation or
Microsoft® Windows NT®Server version 4.0 use DNS name resolution when a
name query contains a name that contains a period or is greater than 15 bytes in
length. Computers running Windows 2000 always try DNS name resolution. For more
 Chapter 5 Introduction to DNS 307

information about DNS and NetBIOS name resolution, see “TCP/IP Troubleshooting”
and “Windows 2000 DNS” in this book.

DNS clients can make two types of queries: recursive and iterative.

Recursive and Iterative Queries

With a recursive name query, the DNS client requires that the DNS server respond to
the client with either the requested resource record or an error message stating that the
record or domain name does not exist. The DNS server cannot just refer the DNS
client to a different DNS server.
Thus, if a DNS server does not have the requested information when it receives a
recursive query, it queries other servers until it gets the information, or until the name
query fails.
Recursive name queries are generally made by a DNS client to a DNS server, or by a
DNS server that is configured to pass unresolved name queries to another DNS server,
in the case of a DNS server configured to use a forwarder.
An iterative name query is one in which a DNS client allows the DNS server to return
the best answer it can give based on its cache or zone data. If the queried DNS server
does not have an exact match for the queried name, the best possible information it
can return is a referral (that is, a pointer to a DNS server authoritative for a lower
level of the domain namespace). The DNS client can then query the DNS server for
which it obtained a referral. It continues this process until it locates a DNS server that
is authoritative for the queried name, or until an error or time-out condition is met.
This process is sometimes referred to as “walking the tree,” and this type of query is
typically initiated by a DNS server that attempts to resolve a recursive name query for
a DNS client.
Figure 5.4 shows an example of iterative and recursive queries. This example assumes
that none of the servers have the requested information in their caches.
308 Part 2 Address Allocation and Name Resolution

Iterative queries

“ ”
Name Server
NameServer1

com
Name Server

reskit.com
Name Server

Recursive query

Resolver

Figure 5.4 Iterative and Recursive Queries

In the example shown in Figure 5.4, a client somewhere on the Internet needs the IP
address of noam.reskit.com. The following events take place:
1. The client contacts NameServer1 with a recursive query for noam.reskit.com.
The server must now return either the answer or an error message.
2. NameServer1 checks its cache and zones for the answer, but does not find it,
so it contacts a server authoritative for the Internet (that is, a root server) with
an iterative query for noam.reskit.com.
3. The server at the root of the Internet does not know the answer, so it responds
with a referral to a server authoritative for the .com domain.
4. NameServer1 contacts a server authoritative for the .com domain with an
iterative query for noam.reskit.com.
5. The server authoritative for the .com domain does not know the exact answer,
so it responds with a referral to a server authoritative for the reskit.com
domain.
6. NameServer1 contacts the server authoritative for the reskit.com domain with
an iterative query for noam.reskit.com.
7. The server authoritative for the reskit.com domain does know the answer. It
responds with the requested IP address.
 Chapter 5 Introduction to DNS 309

8. NameServer1 responds to the client query with the IP address for

noam.reskit.com.

Caching and Time to Live

When a server is processing a recursive query, it might be required to send out several
queries to find the definitive answer. The server caches all of the information that it
receives during this process for a time that is specified in the returned data. This
amount of time is referred to as the time to live (TTL) and is specified in seconds. The
server administrator of the primary zone that contains the data decides on the TTL for
the data. Smaller TTL values help ensure that information about the domain is more
consistent across the network, in the event that this data changes often. However, this
also increases the load on the name servers that contain the name, and it also increases
Internet traffic. Because data is cached, changes made in resource records might not
be immediately available to the entire Internet.
Once data is cached by a DNS server, the DNS server must start decreasing the TTL
from its original value so that it will know when to flush the data from its cache.
When the DNS server answers a query with its cached data, it includes the remaining
TTL for the data. The resolver can then cache this data, using the TTL sent by the
server.

Negative Caching
In addition to caching resolved queries, resolvers and servers can also cache negative
responses, that is, the information that a specific resource record set (RRset) or DNS
domain name does not exist. Negative caching can reduce the response time for
negative answers. It can also reduce network traffic by reducing the number of
messages that must be sent between resolvers and name servers or between name
servers.
Negative caching is specified in RFCs 1034 and 2308. RFC 1034 describes how to
cache negative responses and makes negative caching optional. RFC 2308 requires
resolvers to cache negative responses if they cache any responses. It also describes a
way for name servers to forward cached negative responses to resolvers. Just as with
ordinary caching, they must also start decreasing the TTL.
For more information about negative caching with Windows 2000 DNS, see
“Windows 2000 DNS” in this book.

Resource Records and Zones

To resolve names, servers consult their zones (also called DNS database files or
simply db files). The zones contain resource records (RRs) that make up the resource
information associated with the DNS domain. For example, some resource records
map friendly names to IP addresses, and others map IP addresses to friendly names.
Certain resource records not only include information about servers in the DNS
domain, but also serve to define the domain by specifying which servers are
authoritative for which zones. These resource records, the SOA and NS resource
records, are described in more detail later in this section.
310 Part 2 Address Allocation and Name Resolution

Resource Record Format

Resource records have the following syntax:
Owner TTL Class Type RDATA

Table 5.2 describes each of these fields.

Table 5.2 Typical Resource Record Fields
Name Description
Owner The name of the host or the DNS domain to which this resource record
belongs.
Time to Live A 32-bit integer that represents, in seconds, the length of time that a DNS
server or resolver should cache this entry before it is discarded. This field
is optional, and if it is not specified, the client uses the Minimum TTL in
the SOA record.
Class Defines the protocol family in use. It is almost always IN for the Internet
system. The other value defined in RFC 1034 is CH for the Chaos
system, which was used experimentally at the Massachusetts Institute of
Technology.
Type Identifies the type of resource record.
RDATA The resource record data. It is a variable type that represents the
information being described by the type. For example, in an A record,
this is the 32-bit IP address that represents the host defined by the
resource record.

Resource records are represented in binary form in packets when lookups and
responses are made using DNS. In the database files, however, resource records are
represented as text entries. Most resource records are represented as single-line text
entries. If an entry is going to span more than one line, you can use parentheses to
encapsulate the information. In many implementations of DNS, only the Start of
Authority (SOA) record can be multiple lines. For readability, blank lines and
comments are often inserted in the zone files, and are ignored by the DNS server.
Comments always start with a semicolon (;) and end with a carriage return.

Resource Record Types

Different types of resource records can be used to provide DNS-based data about
computers on a TCP/IP network. This section describes the following resource
records:
• SOA
• NS
• A
• PTR
• CNAME
 Chapter 5 Introduction to DNS 311

• MX
• SRV

Next, it lists some of the other resource records specified by RFC standards. Finally, it
lists resource records that are specific to the Windows 2000 implementation and one
resource record specified by the ATM Forum.
SOA Resource Records
Every zone contains a Start of Authority (SOA) resource record at the beginning of
the zone. SOA resource records include the following fields:
• The Owner, TTL, Class, and Type fields, as described in “Resource Record
Format” earlier in this chapter.
• The authoritative server field shows the primary DNS server authoritative for
the zone.
• The responsible person field shows the e-mail address of the administrator
responsible for the zone. It uses a period (.) instead of an at symbol (@).
• The serial number field shows how many times the zone has been updated.
When a zone’s secondary server contacts the master server for that zone to
determine whether it needs to initiate a zone transfer, the zone’s secondary
server compares its own serial number with that of the master. If the serial
number of the master is higher, the secondary server initiates a zone transfer.
• The refresh field shows how often the secondary server for the zone checks to
see whether the zone has been changed.
• The retry field shows how long after sending a zone transfer request the
secondary server for the zone waits for a response from the master server
before retrying.
• The expire field shows how long after the previous zone transfer the
secondary server for the zone continues to respond to queries for the zone
before discarding its own zone as invalid.
• The minimum TTL field applies to all the resource records in the zone
whenever a time to live value is not specified in a resource record. Whenever a
resolver queries the server, the server sends back resource records along with
the minimum time to live. Negative responses are cached for the minimum
TTL of the SOA resource record of the authoritative zone.
The following example shows the SOA resource record:
noam.reskit.com. IN SOA (
noamdc1.noam.reskit.com. ; authoritative server
for the zone
administrator.noam.reskit.com. ; zone admin e-mail
; (responsible person)
5099 ; serial number
3600 ; refresh (1 hour)
600 ; retry (10 mins)
312 Part 2 Address Allocation and Name Resolution

86400 ; expire (1 day)

60 ) ; minimum TTL (1 min)

NS Resource Records
The name server (NS) resource record indicates the servers authoritative for the zone.
They indicate primary and secondary servers for the zone specified in the SOA
resource record, and they indicate the servers for any delegated zones. Every zone
must contain at least one NS record at the zone root.
For example, when the administrator on reskit.com delegated authority for the
noam.reskit.com subdomain to noamdc1.noam.reskit.com., the following line was
added to the zones reskit.com and noam.reskit.com:
noam.reskit.com. IN NS noamdc1.noam.reskit.com.

A Resource Records
The address (A) resource record maps an FQDN to an IP address, so the resolvers can
request the corresponding IP address for an FQDN. For example, the following A
resource record, located in the zone noam.reskit.com, maps the FQDN of the server to
its IP address:
noamdc1 IN A 172.16.48.1

PTR Records
The pointer (PTR) resource record, in contrast to the A resource record, maps an IP
address to an FQDN. For example, the following PTR resource record maps the IP
address of noamdc1.noam.reskit.com to its FQDN:
1.48.16.172.in-addr.arpa. IN PTR
noamdc1.noam.reskit.com.

CNAME Resource Records

The canonical name (CNAME) resource record creates an alias (synonymous name)
for the specified FQDN. You can use CNAME records to hide the implementation
details of your network from the clients that connect to it. For example, suppose you
want to put an FTP server named ftp1.noam.reskit.com on your noam.reskit.com
subdomain, but you know that in six months you will move it to a computer named
ftp2.noam.reskit.com, and you do not want your users to have to know about the
change. You can just create an alias called ftp.noam.reskit.com that points to
ftp1.noam.reskit.com, and then when you move your computer, you need only change
the CNAME record to point to ftp2.noam.reskit.com. For example, the following
CNAME resource record creates an alias for ftp1.noam.reskit.com:
ftp.noam.reskit.com. IN CNAME ftp1.noam.reskit.com.

Once a DNS client queries for the A resource record for ftp.noam.reskit.com, the DNS
server finds the CNAME resource record, resolves the query for the A resource record
for ftp1.noam.reskit.com, and returns both the A and CNAME resource records to the
client.
 Chapter 5 Introduction to DNS 313

Note According to RFC 2181, there must be only one canonical name per alias.

MX Resource Records
The mail exchange (MX) resource record specifies a mail exchange server for a DNS
domain name. A mail exchange server is a host that will either process or forward
mail for the DNS domain name. Processing the mail means either delivering it to the
addressee or passing it to a different type of mail transport. Forwarding the mail
means sending it to its final destination server, sending it using Simple Mail Transfer
Protocol (SMTP) to another mail exchange server that is closer to the final
destination, or queuing it for a specified amount of time.
Note Only mail exchange servers use MX records.

If you want to use multiple mail exchange servers in one DNS domain, you can have
multiple MX resource records for that domain. The following example shows MX
resource records for the mail servers for the domain noam.reskit.com.:
*.noam.reskit.com. IN MX 0 mailserver1.noam.reskit.com.
*.noam.reskit.com. IN MX 10 mailserver2.noam.reskit.com.
*.noam.reskit.com. IN MX 10 mailserver3.noam.reskit.com.

The first three fields in this resource record are the standard owner, class, and type
fields. The fourth field is the mail server priority, or preference value. The preference
value specifies the preference given to the MX record among MX records. Lower
priority records are preferred. Thus, when a mailer needs to send mail to a certain
DNS domain, it first contacts a DNS server for that domain and retrieves all the MX
records. It then contacts the mailer with the lowest preference value.
For example, suppose Jane Doe sends an e-mail message to
[email protected] on a day that mailserver1 is down, but mailserver2 is
working. Her mailer tries to deliver the message to mailserver1, because it has the
lowest preference value, but it fails because mailserver1 is down. This time, Jane’s
mailer can choose either mailserver2 or mailserver3, because their preference values
are equal. It successfully delivers the message to mailserver2.
To prevent mail loops, if the mailer is on a host that is listed as an MX for the
destination host, the mailer can deliver only to an MX with a lower preference value
than its own host.
Note The sendmail program requires special configuration if a CNAME is not
referenced in the MX record.

SRV Records
With MX records, you can have multiple mail servers in a DNS domain, and when a
mailer needs to send mail to a host in the domain, it can find the location of a mail
exchange server. But what about other applications, such as the World Wide Web or
telnet?
314 Part 2 Address Allocation and Name Resolution

Service (SRV) resource records enable you to specify the location of the servers for a
specific service, protocol, and DNS domain. Thus, if you have two Web servers in
your domain, you can create SRV resource records specifying which hosts serve as
Web servers, and resolvers can then retrieve all the SRV resource records for the Web
servers.
The format of an SRV record is as follows:
_Service._Proto.Name TTL Class SRV Priority Weight Port Target
• The _Service field specifies the name of the service, such as http or telnet.
Some services are defined in the standards, and others can be defined locally.
• The _Proto field specifies the protocol, such as TCP or UDP.
• The Name field specifies the domain name to which the resource record refers.
• The TTL and Class fields are the same as the fields defined earlier in this
chapter.
• The Priority field specifies the priority of the host. Clients attempt to contact
the host with the lowest priority.
• The Weight field is a load balancing mechanism. When the priority field is the
same for two or more records in the same domain, clients should try records
with higher weights more often, unless the clients support some other load
balancing mechanism.
• The Port field shows the port of the service on this host.
• The Target field shows the fully qualified domain name for the host
supporting the service.

The following example shows SRV records for Web servers:

_http._tcp.reskit.com. IN SRV 0 0 80 webserver1.noam.reskit.com.
_http._tcp.reskit.com. IN SRV 10 0 80 webserver2.noam.reskit.com.

Note This example does not specify a TTL. Therefore, the resolver uses the
minimum TTL specified in the SOA resource record.

If a computer needs to locate a Web server in the reskit.com DNS domain, the
resolver sends the following query:
_http._tcp.www.reskit.com.

The DNS server replies with the SRV records listed above. The resolver then chooses
between WebServer1 and WebServer2 by looking at their priority values. Because
WebServer1 has the lowest priority value, the DNS server chooses WebServer1.
Note If the priority values had been the same, but the weight values had been
different, the client would have chosen a Web server randomly, except that the
server with the highest weight value would have had a higher probability of being
chosen.
 Chapter 5 Introduction to DNS 315

Next, the resolver requests the A record for webserver1.reskit.com, and the DNS
server sends the A record. Finally, the client attempts to contact the Web server.
For more information about SRV records, see the link to the Internet Engineering
Task Force (IETF) on the Web Resources page at
http://windows.microsoft.com/windows2000/reskit/webresources. Windows 2000
supports the Internet Draft titled “A DNS RR for specifying the location of services
(DNS SRV).”
Less Common Resource Records
Table 5.3 shows some other resource records and the RFCs that define them. Many of
these resource records are considered experimental.
Table 5.3 Less Common Resource Record Types
Record Type RFC Description
AAAA 1886 Special address record that maps a host (computer or other
network device) name to an IPv6 address.
AFSDB 1183 Gives the location of either an Andrew File System (AFS)
cell database server, or a Distributed Computing Environment
(DCE) cell's authenticated server. The AFS system uses DNS
to map a DNS domain name to the name of an AFS cell
database server. The Open Software Foundation's DCE
Naming Service uses DNS for a similar function.
HINFO 1035 The host information resource record identifies a host's
hardware type and operating system. The CPU Type and
Operating System identifiers come from the computer names
and system names listed in RFC 1700.
1183 The Integrated Services Digital Network (ISDN) resource
record is a variation of the A (address) resource record.
Rather than mapping an FQDN to an IP address, the ISDN
record maps the name to an ISDN address. An ISDN address
is a phone number that consists of a country/region code, an
area code or country/region code, a local phone number, and
optionally, a subaddress. The ISDN resource record is
designed to be used in conjunction with the route through
(RT) resource record.
MB 1035 The mailbox (MB) resource record is an experimental record
that specifies a DNS host with the specified mailbox. Other
related experimental records are the mail group (MG)
resource record, the mailbox rename (MR) resource record,
and the mailbox information (MINFO) resource record.
MG 1035 The mail group (MG) resource record is an experimental
record that specifies a mailbox that is a member of the mail
group (mailing list) specified by the DNS domain name.
Other related experimental records are the MB resource
record, the MR resource record, and the MINFO resource
record.
MINFO 1035 The MINFO resource record is an experimental record that
316 Part 2 Address Allocation and Name Resolution

specifies a mailbox that is responsible for the specified

mailing list or mailbox. Other related experimental records
are the MB resource record, the MG resource record, and the
MR resource record.
(continued)
Table 5.3 Less Common Resource Record Types (continued)
Record Type RFC Description
MR 1035 The MR resource record is an experimental record that
specifies a mailbox that is the proper rename of another
specified mailbox. Other related experimental records are the
MB resource record, the MG resource record, and the
MINFO resource record.
RP 1183 Identifies the responsible person (RP) for the specified DNS
domain or host.
RT 1183 The route through (RT) resource record specifies an
intermediate host that routes packets to a destination host.
The RT record is used in conjunction with the ISDN and X25
resource records. It is syntactically and semantically similar
to the MX record type and is used in much the same way.
TXT 1035 The text resource (TXT) record associates general textual
information with an item in the DNS database. A typical use
is for identifying a host's location (for example, Location:
Building 26S, Room 2499). A single TXT record can contain
multiple strings, up to 64 kilobytes (KB).
WKS 1035 The well-known service (WKS) resource record describes the
services provided by a particular protocol on a particular
interface. The protocol is usually UDP or TCP, but can be
any of the entries listed in the Windows 2000 Protocols file
located in %SystemRoot%\System32\Drivers\Etc\Protocol.
The services are the services below port number 256 from the
Windows 2000 Services file located
in %SystemRoot%\System32\Drivers\Etc\Services.
X.25 1183 The X.25 resource record is a variation of the A (address)
resource record. Rather than mapping a FQDN to an IP
address, the X.25 record maps the name to an X.121 address.
X.121 is the International Standards Organization (ISO)
standard that specifies the format of addresses used in X.25
networks. The X.25 resource record is designed to be used in
conjunction with the route through (RT) resource record.

Resource Records Not Defined in RFCs

In addition to the resource record types listed in the RFCs, Windows 2000 uses the
following resource record types, shown in Table 5.4.
 Chapter 5 Introduction to DNS 317

Table 5.4 Resource Record Types Not Defined in the RFCs

Name Description
WINS The Windows 2000 DNS server can use a WINS server for looking
up the host portion of a DNS name that does not exist in the DNS
zone authoritative for the name.
WINS reverse This entry is used in a reverse lookup zone for finding the host
lookup (WINS-R) portion of the DNS name if given its IP address. A DNS server
issues a NetBIOS adapter status query if the zone authoritative for
the queried IP address does not contain the record and does contain
the WINS-R resource record.
ATMA The ATMA resource record, defined by the ATM Forum, is used to
map DNS domain names to ATM addresses. For more information,
contact the ATM Forum for the ATM Name System Specification
Version 1.0.

Delegation and Glue Records

Delegation and glue records are records that you add to a zone in order to delegate a
subdomain into a separate zone. A delegation is an NS record in the parent zone that
lists the name server authoritative for the delegated zone. A glue record is an A record
for the name server authoritative for the delegated zone.
For example, suppose the name server for the DNS domain reskit.com, delegated
authority for the noam.reskit.com zone to the name server noamNS.noam.reskit.com.
You add the following records to the reskit.com zone:
noam.reskit.com. IN NS noamNS.noam.reskit.com
noamNS.noam.reskit.com. IN A 172.16.54.1

Delegations are necessary for name resolution. Glue records are also necessary if the
name server authoritative for the delegated zone is also a member of that domain. A
glue record is necessary in the example above because noamNS.noam.reskit.com. is a
member of the delegated domain noam.reskit.com. However, if it was a member of a
different domain, the resolver can perform standard name resolution to resolve the
name of the authoritative name server to an IP address.
When a resolver submits a query for a name in the child zone to the name server that
is authoritative for the parent zone, the server authoritative for the parent zone checks
its zone. The delegation tells it which name server is authoritative for the child zone.
The server authoritative for the parent zone can then return a referral to the resolver.

Zones
The DNS standards do not specify the internal data structure that stores resource
records, and various implementations differ. Generally, servers use zones stored on
that server in plain text, but it is not required. With Windows 2000, you can integrate
your DNS database with the Active Directory database, in which case the zones are
stored in the Active Directory database.
318 Part 2 Address Allocation and Name Resolution

One common implementation of DNS, the Berkeley Internet Name Domain (BIND)
implementation, generally uses the file names shown in Table 5.5.
Table 5.5 Zone Names Used in BIND
Name Description
db.domain Forward lookup zone. For example, if your DNS domain is reskit.com,
then this file is called db.reskit.com.
db.addr Reverse lookup zone. For example, if your network is the class C
network address 172.16.32 then this file is called db.172.16.32.
db.cache Also known as the root hints file, this file contains the names and IP
addresses for the name servers that maintain the root DNS domain. This
file is essentially the same on all servers that use Internet root DNS
servers, but must be modified for servers that use private root DNS
servers. (A root DNS server is a DNS server that is authoritative for the
root of the namespace.)
db.127.0.0.1 Used to resolve queries to the loopback address. It is essentially the same
on all name servers.

The names of the database files are arbitrary and are specified in the configuration of
the DNS server. By default, the Microsoft Windows 2000 DNS server does not use
the same file names as a typical BIND DNS server but instead uses zone_name.dns.
However, if you are porting DNS db files from another DNS server, you can
configure the Microsoft Windows 2000 DNS server to use the BIND file names.
The following sections explain the contents of the zones and describe one additional
file, the BOOT file, which is used by BIND servers, though not specified in the DNS
standards.
Forward Lookup Zone
Forward lookup zones contain information needed to resolve names within the DNS
domain. They must include SOA and NS records and can include any type of resource
record except the PTR resource record.
Reverse Lookup Zone
Reverse lookup zones contain information needed to perform reverse lookups. They
usually include SOA, NS, PTR, and CNAME records.
With most queries, the client supplies a name and requests the IP address that
corresponds to that name. This type of query is typically described as a forward
lookup.
But what if a client already has a computer’s IP address and wants to determine the
DNS name for the computer? This is important for programs that implement security
based on the connecting FQDN, and is also used for TCP/IP network troubleshooting.
The DNS standard provides for this possibility through reverse lookups.
If the only means to answer a reverse lookup were to conduct a thorough search of all
DNS domains in the DNS namespace, the reverse query search would be too
exhaustive to perform in any practical way.
 Chapter 5 Introduction to DNS 319

To solve this problem, a special DNS domain called in-addr.arpa was created. This
domain uses a reverse ordering of the numbers in the dotted-decimal notation of IP
addresses. With this arrangement, administration of lower limbs of the in-addr.arpa
domain can be delegated to organizations as they are assigned their class A, B, or C IP
network IDs. For more information about creating classless reverse lookup zones, see
“Windows 2000 DNS” in this chapter. See also RFC 2317, “Classless IN-
ADDR.ARPA delegation.”
Figure 5.5 shows a branch of the in-addr.arpa namespace.
“ ” (Root)

arpa

in-addr

172

16

44

Figure 5.5 In-addr.arpa Namespace

The in-addr.arpa domain tree requires PTR resource records to store and provide
reverse mappings for IP addresses of their corresponding FQDNs.
If a client needs to find the FQDN associated with the IP address 172.16.44.1, the
client queries for the PTR record of the 1.44.16.172.in-addr.arpa domain name.
320 Part 2 Address Allocation and Name Resolution

Inverse Queries
In addition to reverse lookups, some DNS servers support what is known as
an inverse query. Just as with a reverse lookup, a client making an inverse
query provides the IP address and requests the FQDN. However, the server
does not use the in-addr.arpa domain to find the answer, and it does not
query any other servers. Instead, it simply checks its own zones for the
answer, and if it does not find the answer, it returns an error message.
There is no way for either the server or the client to know whether the IP
address is simply missing from the zones of that server, or whether the IP
address does not exist.
Because support for inverse queries is optional and because servers often
cannot provide a definitive answer, inverse queries are of limited use. Only
certain applications use inverse queries, such as earlier versions of
nslookup.
The Windows 2000 server responds to inverse query requests by replying
with the IP address specified in the query enclosed in square brackets. For
example, if it receives an inverse query for 172.16.72.1, it responds with
[172.16.72.1].
For more information about inverse queries, see RFC 1035.

Root Hints Files

The root hints file, also called the cache hints file, contains host information that is
needed to resolve names outside of the authoritative DNS domains. It contains the
names and addresses of root DNS servers.
If your network is connected to the Internet, the root hints file should contain records
for the root DNS servers on the Internet. In Windows 2000, a root hints file with
current Internet root DNS mappings is installed as the file Cache.dns in the
directory %SystemRoot%\System32\Dns.
If your network is not connected to the Internet, you must replace the NS and A
records in the cache file with NS and A records for the DNS servers that are
authoritative for the root of your private TCP/IP network.
For example, suppose you have created two internal root DNS servers
(InternalRoot1.reskit.com. and InternalRoot2.reskit.com.). You then create a root hints
file on other DNS servers in your network, pointing to the internal root DNS servers.
That way, if other name servers receive a query they cannot resolve, they can simply
query the internal root servers specified in the file.
The following reskit root hints file provides NS resource records and name-to-IP
address mappings for name servers in the reskit.com domain.
; Internal root hints file for reskit.com, which is not connected to
; the Internet
 Chapter 5 Introduction to DNS 321

. 86400 IN NS InternalRoot1.reskit.com.
. 86400 IN NS InternalRoot2.reskit.com.

InternalRoot1.reskit.com. 86400 IN NS 172.16.64.1

InternalRoot2.reskit.com. 86400 IN NS 172.16.64.2

Note The Windows 2000 Configure DNS Server wizard makes a best effort to
determine whether the network is connected to the Internet and, if not, creates
its own cache file.

Boot Files
Although the boot file is not actually defined in the RFCs and is not needed for a
server to be RFC compliant, it is described here for completeness. This file is part of
the BIND-specific implementation of DNS.
It is not used by default in the Microsoft implementation of DNS. If advantageous,
you can port an existing BIND boot file to a Microsoft DNS server. The
Windows 2000 DNS server supports only a subset of the BIND boot file directives,
and only for the file type used by BIND 4.x.x servers.
The BIND Boot file controls the startup behavior of the DNS server. Commands must
start at the beginning of a line and no spaces can precede commands. Table 5.6 shows
descriptions of some of the boot file commands supported by Windows 2000.
Table 5.6 Descriptions of Boot File Commands
Command Description
Directory command Specifies a directory where other files referred to in the boot file
can be found.
Cache command Specifies a file used to help the DNS server contact name servers
for the root domain. This command and the file it refers to must
be present.
(continued)
Table 5.6 Descriptions of Boot File Commands (continued)
Command Description
Primary command Specifies a primary zone for which this name server is
authoritative and a zone file that contains the resource records for
that zone. Multiple primary command records can exist in the
boot file.
Secondary command Specifies a secondary zone for which this name server is
authoritative and a list of IP addresses of master servers from
which to attempt zone transfer. It also defines the name of the
local file for saving this zone. Multiple secondary command
records can exist in the boot file.
322 Part 2 Address Allocation and Name Resolution

Table 5.7 shows the syntax of some of the boot file commands supported by
Windows 2000.
Table 5.7 Syntax of Boot File Commands Supported by Windows 2000
Syntax Example
directory <directory> directory c:\winnts\system32\dns
cache <file name> cache cache
primary <domain> <file primary reskit.com reskit.com.dns
name> primary dev.reskit.com dev.reskit.com.dns
secondary <domain> secondary test.reskit.com 157.55.200.100
<hostlist> <local file name> test.reskit.com.dns
forwarders <hostlist> forwarders 172.16.64.4 172.16.64.8
slave (follows the forwarders forwarders 172.16.64.4 172.16.64.8 slave
parameter)

For more information about the BIND boot file, see the link to the Microsoft TechNet
Web site on the Web Resources page at
http://windows.microsoft.com/windows2000/reskit/webresources. On the TechNet
Web site, search for the keywords “structure” and “domain name system” and “boot
file.”

Zone Transfer
When changes are made to the zone on a master server, these changes must be
replicated to all the secondary servers for that zone, using a mechanism called zone
transfer. In the original DNS specifications, only one form of zone transfer was
available, known as full zone transfer. New RFCs discuss an additional type of zone
transfer: incremental zone transfer. This section describes both types. It also describes
DNS Notify, a mechanism by which the master server for a zone can notify the
secondary servers of zone changes.

Full Zone Transfer

In a full zone transfer, defined in the original DNS specifications, the master server
for a zone transmits the entire zone database to the secondary server for that zone.
Secondary servers initiate full zone transfers using the following process:
1. The secondary server for the zone waits a certain amount of time (specified in
the Refresh field of the SOA resource record), and then polls the master server
for its SOA.
2. The master server for the zone responds with the SOA resource record.
3. The secondary server for the zone compares the returned serial number to its
own serial number. If the serial number sent by the master server for the zone
is higher than its own serial number, that means its zone database is out of
date, and it sends an AXFR request (a request for a full zone transfer).
 Chapter 5 Introduction to DNS 323

4. The master server for the zone sends the full zone database to the secondary
server.

If the master server for the zone does not respond to polling by the secondary server,
the secondary server continues to retry after the interval specified in the Retry field of
the SOA resource record. If there is still no answer after the interval specified in the
Expire field since the last successful zone transfer, it discards its zone.
Note Name servers running versions of BIND earlier than 4.9.4 can send and
receive only one resource record per message during a full zone transfer.
Name servers running versions of BIND 4.9.4 and later, and Windows 2000, can send
and receive multiple resource records per message. This improves the performance of
full zone transfers.
However, to provide backward compatibility, name servers running Windows 2000
default to sending only one resource record per message if any secondary servers that
are not running Windows are configured for the zone. If you have secondary name
servers running versions of BIND 4.9.4 and later, configuring Windows 2000 to send
multiple resource records per message improves performance. For more information,
see the link to the Microsoft TechNet Web site on the Web Resources page at
http://windows.microsoft.com/windows2000/reskit/webresources. On the TechNet
Web site, search for “DNS Compatibility,” “BIND,” and “4.9.4.”

Incremental Transfer
Full zone transfers can consume a great deal of network bandwidth, especially for
complex DNS configurations. To solve this problem, RFC 1995 specifies an
additional standard, incremental zone transfer. With incremental zone transfer, only
the modified part of the zone must be transferred.
Incremental zone transfer works much the same as full zone transfer. The secondary
server for the zone still uses the SOA resource record to determine when to poll the
master server for the zone, when to retry, and so on. However, if it needs to perform a
zone transfer, it sends an incremental zone transfer (IXFR) query instead of an AXFR
query, requesting that the master server for the zone perform an incremental zone
transfer.
The master server for the zone, meanwhile, maintains a recent version history of the
zone, which observes any record changes that occurred in the most recent version
updates of the zone. Then, if the master server for the zone has a newer version of the
zone, it can forward only those record changes that have occurred between the two
different versions of the zone (the current versions on the master and secondary
servers) to the secondary server for the zone. The master server sends the oldest
updates first and the newest updates last.
When the secondary server receives an incremental zone transfer, it creates a new
version of the zone and begins replacing its resource records with the updated
resource records, starting with the oldest updates and ending with the newest updates.
When all of the updates have been made, the secondary server replaces its old version
of the zone with the new version of the zone.
324 Part 2 Address Allocation and Name Resolution

The master server for the zone is not required to perform an incremental transfer. It
can perform a full zone transfer if it does not support incremental zone transfer, if it
does not have all the necessary data for performing an incremental zone transfer, or if
an incremental zone transfer takes more bandwidth than a full zone transfer.
Even though incremental zone transfer saves network bandwidth, it uses space on the
server to record the version history. To conserve space, servers can purge the version
history.

DNS Notify
DNS Notify is a revision to the DNS standard (RFC 1996) that proposes that the
master server for a zone notify certain secondary servers in that zone of changes, and
the secondary servers can then check to see whether they need to initiate a zone
transfer. This process can help improve consistency of zone data among secondary
servers.
To determine which secondary servers in a zone to send the changes to, the master
server for the zone contains a notify list, which is a list of the IP addresses for those
secondary servers. The master server for the zone notifies only listed servers when
zone updates occur.
When the local zone on a master server for the zone is updated, the following events
take place:
1. The Serial Number field in the SOA record is updated to indicate that a new
version of the zone has been written to a disk.
2. The master server then sends a notify message to other servers that are part of
its notify list.
3. All secondary servers for the zone that receive the notify message respond by
initiating an SOA-type query back to the notifying master server to determine
if the zone of the notifying server is a later version than the currently stored
copy of the zone.
4. If a notified server determines that the serial number used in the SOA record of
the zone of the notifying server is higher (more recent) than the serial number
used in the SOA record for its current zone copy, the notified server requests
an AXFR or IXFR zone transfer.

Note On servers running Windows 2000, you can configure the Notify set.

Dynamic Update
Dynamic update is a new standard, specified in RFC 2136, that provides a means of
dynamically updating zone data on a zone’s primary server.
Originally, DNS was designed to support only static changes to a zone database.
Because of the design limitations of static DNS, the ability to add, remove, or modify
resource records could only be performed manually by a DNS system administrator.
For example, a DNS system administrator would edit records on a zone’s primary
server and the revised zone database is then propagated to secondary servers during
 Chapter 5 Introduction to DNS 325

zone transfer. This design is workable when the number of changes is small and
updates occur infrequently, but can otherwise become unmanageable.
With dynamic update, on the other hand, the primary server for the zone can also be
configured to support updates that are initiated by another computer or device that
supports dynamic update. For example, it can receive updates from workstations
registering A and PTR resource records, or from DHCP servers. Updates are sent
using a standard UPDATE message format and can include the addition or deletion of
individual resource records (RRs) or sets of resource records (RRsets). For
information about the UPDATE message format, see RFC 2136.
In order for a request for a dynamic update to be performed, several prerequisite
conditions can also be identified. Where prerequisites are set, all such conditions must
be met before an update is allowed. Some examples of prerequisites that can be set
are:
• A required RR or RRset already exists or is in use prior to an update.
• A required RR or RRset does not exist or is not in use prior to an update.
• A requester is permitted to initiate an update of a specified RR or RRset.

Each prerequisite must be satisfied in order for an update to occur. After all
prerequisites are met, the zone’s primary server can then proceed with an update of its
local zones. Multiple updates can be processed concurrently only if one update does
not depend on the final result of another update.

DNS Standards
Although the core DNS standards (as set forth in RFCs 1034 and 1035) have been
well-established since their acceptance as standards in 1987, numerous revisions to
DNS have been made in the intervening years. These revisions have been brought
about by additional RFCs and Internet drafts, which are independently authored and
submitted to the IETF for further circulation and review before being accepted as
standards throughout the Internet.
Table 5.8 lists the accepted and proposed RFC standards that Windows 2000 supports.
Table 5.8 DNS-related RFC Standards Supported by Windows 2000
Number Status Title
1034 Standard Domain Names - Concepts and Facilities
1035 Standard Domain Names - Implementation and Specification
1123 Standard Requirements for Internet Hosts - Application and
Support
1886 Proposed DNS Extensions to Support IP Version 6
1995 Proposed Incremental Zone Transfer in DNS
1996 Proposed A Mechanism for Prompt DNS Notification of Zone
Changes
2136 Proposed Dynamic Updates in the Domain Name System (DNS
UPDATE)
326 Part 2 Address Allocation and Name Resolution

2181 Standards Track Clarifications to the DNS Specification

2308 Standards Track Negative Caching of DNS Queries (DNS NCACHE)

Additional Resources
• For more information about conceptual information about DNS, see DNS and
BIND, 3rd Edition by Paul Albitz and Cricket Liu, 1998, O’Reilley &
Associates, Inc..
• For more information about Requests for Comments (RFCs) and Internet
drafts, see the link to the IETF on the Web Resources page at
http://windows.microsoft.com/windows2000/reskit/webresources.