To understand how DNS works, imagine a typical user, John, logging in to his computer. He launches his web browser so that he can view the website at a company, ExampleCo (see the image below). He enters the name of their website—http://www.example.com. Then:

1. John’s workstation sends a request to the DNS server about the IP address of www.example.com.
2. The DNS server checks its database to find that www.example.com corresponds to 192.168.1.4.
3. The server returns this address to John’s browser.
4. The browser uses the address to locate the website.
5. The browser displays the website on John’s monitor.

This section provides an overview of the concepts in DNS.

John can access the ExampleCo website because his DNS server knows the www.example.com IP address. The server learned the address by searching through the domain namespace. DNS was designed as a tree structure, where each named domain is a node in the tree. The top-most node of the tree is the DNS root domain (.), under which there are subdomains, such as .com, .edu, .gov, and .mil (see the image below).

The fully qualified domain name (FQDN) is a dot-separated string of all the network domains leading back to the root. This name is unique for each host on the Internet. The FQDN for the sample domain is example.com., with its domain example, parent domain .com, and root domain “.” (dot).

DNS is based on a client/server model. In this model, nameservers store data about a portion of the DNS database and provide it to clients that query the nameserver across the network. Nameservers are programs that run on a physical host and store zone data. As administrator for a domain, you set up a nameserver with the database of all the resource records (RRs) describing the hosts in your zone or zones (see the image below).

The DNS servers provide name-to-address translation, or name resolution. They interpret the information in a fully qualified domain name (FQDN) to find its address.

Each zone must have one primary nameserver that loads the zone contents from a local database, and a number of secondary servers, which load a copy of the data from the primary server (see the image below). This process of updating the secondary server from the primary server is called a zone transfer.

Even though a secondary nameserver acts as a kind of backup to a primary server, both types of servers are authoritative for the zone. They both learn about hostnames in the zone from the zone authoritative database, not from information learned while answering queries. Clients can query both servers for name resolution.

As you configure the Cisco Prime Network Registrar DNS nameserver, you specify what role you want the server to perform for a zone—primary, secondary, or caching-only. The type of server is meaningful only in context to its role. An authoritative DNS server can only be a primary or a secondary server for a zone, it does not specify zones for caching servers.

In Cisco Prime Network Registrar, the authoritative and caching services are separated and are handled by two separate servers. The authoritative server holds authoritative zone data and responds only to queries for which it is authoritative. The caching server is the recursive/caching server and does not contain any authoritative zone data.

To configure the:

• Primary nameserver, see Managing Primary DNS Servers.
• Secondary nameserver, see Managing Secondary Servers.

The DNS servers described so far perform name-to-address resolution. They can do this easily by searching through their database for the correct address, because they index all the data by name. However, there are times when you need address-to-name resolution so that you can interpret certain output, such as computer log files.

Finding a domain name when you only know the address, however, would require searching the entire namespace. DNS solves this problem by supporting a domain namespace that uses addresses as names, known as the in-addr.arpa or .arpa domain. This reverse zone contains subdomains for each network based on the network number. For consistency and natural grouping, the four octets of a host number are reversed.

The IP address as a domain name appears backward, because the name is in leaf-to-root order. For example, the ExampleCo example domain network number is 192.168.1.0. Its reverse zone is 1.168.192.in-addr.arpa. If you only know the DNS server address (192.168.1.1), the query to the reverse domain would find the host entry 1.1.168.192.in-addr.arpa that maps back to example.com.

Reverse domains are handled through Pointer (PTR) RRs, as indicated in the image below.

Starting from release 8.0, the DNS server functionality is enhanced to provide separate DNS servers for authorization and caching. With this enhancement, Cisco Prime Network Registrar supports DNS64, DNSSEC, Domain Redirect, full IPv6, and has improved caching performance.

Because there can be only one primary DNS server per zone, failure of this server makes it impossible to update the zone data. These updates can occur on the primary DNS server only; software such as DHCP servers, that update DNS resource records must send the updates directly to the primary. A second primary server can become a hot standby that shadows the main primary. This is called High-Availability (HA) DNS.

To send a DNS message above 512 bytes over UDP, you need to use an extension of the DNS protocol known as Extended DNS (EDNS). The EDNS protocol expands the number of flags, label types, and return codes available to the DNS protocol. A version of EDNS specified by RFC 6891 is known as EDNS0. EDNS uses a pseudo resource record known as OPT Resource Record (OPT RR). OPT RR differentiates conventional DNS from EDNS. OPT RRs appear only in the route transmission between DNS clients and servers, they are not cached or persisted to disk. A DNS endpoint that marks a DNS packet as EDNS must insert an OPT RR in the additional data section of the DNS request or response.

The Authoritative and the Caching DNS servers support the EDNS0 extension, but no option codes. You can modify the UDP payload size of the DNS server. The minimum UDP payload size of the DNS server is 512 bytes. The maximum UDP packet size is 64 KB, the default and recommended size for the Caching server is 4KB.

Note

The DNS Server can handle requests from clients that do not support EDNS0, however, the DNS server is not permitted to use any extended capabilities, when it handles requests from clients that do not support EDNS0. The response to client requests are inserted into a default 512 byte message. Clients may indicate that they support EDNS by including an OPT RR in the query. If a server does not support EDNS (or the support is disabled), the server will return FORMERR and the client retries without EDNS. If an answer is larger than the size that the client has reported (either with EDNS or the default 512 bytes), the server will mark the result as truncated and the client may retry using TCP.

DNS views allows you to present alternate versions of zone data to different communities of clients using a single name server.

For example, a DNS server for example.com could maintain two views of the zone, where the view of example.com that can be queried internally includes many hosts that do not exist in the external view. Each zone view is treated as an independent copy of the zone. The DNS server, when answering queries on the zone, uses the match criteria defined in each view to determine the matching zone for the client. The query is then answered based on that zone contents. In some cases, the zone contents may only vary slightly between views.

Note	Cisco Prime Network Registrar 8.2 and later support the DNS views.