The Internet’s naming system is distributed, not one giant address book.

The DNS hierarchy divides administrative responsibility. Root information leads toward top-level domains, and delegations lead toward the servers responsible for particular domain zones. Different organizations can maintain their own data without requiring one operator to edit every Internet name.

Resolvers combine that distributed structure with caching. A recursive resolver can look up information on a client's behalf, then temporarily reuse an answer according to its caching lifetime. Most everyday queries therefore do not require a fresh trip through the complete hierarchy.

DNS stores more than address mappings. Record types can identify mail servers, publish text information, or point one name toward another. Thinking of DNS only as a telephone book misses its role as a general-purpose system for publishing structured information under domain names.

Before DNS, Internet naming relied heavily on a centrally maintained hosts file that participants copied. As the network grew, coordinating updates and distributing that list became increasingly cumbersome. A mechanism allowing different administrators to maintain different parts of the namespace offered a more scalable approach.

Paul Mockapetris introduced DNS in 1983. The foundational specifications evolved, with RFC 1034 describing concepts and facilities and RFC 1035 describing implementation details in 1987. These documents established the distributed hierarchy and message structures underlying a system that has since gained many extensions.

A domain name and a DNS zone are related but not identical concepts. A zone is an administrative portion of the namespace served authoritatively. A parent can delegate a child portion to other authoritative servers, creating a boundary between responsibilities while keeping the overall naming hierarchy connected.

Consider a lookup for a website. A client usually asks a configured recursive resolver. If no suitable cached answer exists, that resolver can consult the hierarchy, following referrals until it reaches an authoritative server for the relevant data. The client does not normally perform every one of those steps itself.

Caching improves speed and reduces repeated work, but it explains why changes can appear at different times for different users. Previously cached records may remain usable until their time to live expires. There is no single instant when every resolver worldwide must discard every older answer after an administrator makes an update.

Distributed does not mean that every part has equal authority, nor does it mean the system has no critical dependencies. DNS can be affected by server failures, configuration errors, attacks, or resolver policies. Redundant authoritative servers and careful operations improve resilience, while DNSSEC can authenticate signed DNS data without encrypting ordinary queries.

DNS connects human-readable naming with the information applications need to locate services. Symbolics.com provides a registration milestone, while IPv6 illustrates one kind of address a name can lead to. Neither story implies that one central machine stores every current answer or that a domain must identify one permanent computer.