Domain Name System

DNS organizes names hierarchically and delegates parts of the namespace to different administrators. A name can have records of several types, supporting addresses, mail, and other uses. This distributes authoritative information instead of requiring every organization to edit one universally shared host table containing all names and their associated data.

References: [1]

Applications usually ask a resolver, which uses existing knowledge to contact relevant servers. A server may answer authoritatively for its zone or refer the requester toward another source. Recursive and iterative lookup allocate responsibility differently, making it important to identify which component continues the search on behalf of a client.

References: [2]

Record types determine how DNS data is interpreted; address records are only one example. Caches retain answers for bounded lifetimes to reduce repeated queries. After an authoritative change, existing caches can persist until their rules require renewal, so publication of a change and its observation by different resolvers are distinct events.

References: [1]

Lookup supplies information needed for subsequent communication, which still requires connections, application protocols, and appropriate identity checks. Names can map to several addresses, and addresses can change as services move. Separating naming from routing and identity explains how a stable service name can outlast changes to its servers and network locations.

References: [2]