DNS Components

DNS separates the client-facing recursion role from authoritative publication. Understanding that boundary prevents a cached answer from being mistaken for the owner of a zone.

Stub and Recursive Resolvers

A stub resolver in an application or operating system sends queries to a configured recursive resolver. The recursive resolver returns a final answer, error, or referral outcome after using cache and, when necessary, performing iterative queries. Its reply can carry the authoritative-answer flag only when the answering server is authoritative for the data; recursion alone does not make it authoritative.

CheckWhat does a recursive resolver do for a stub client?

Obtains a final DNS result using cache and other name servers. Replaces every network router on the packet path. Becomes authoritative for every record it caches.

Authoritative Name Servers

An authoritative server answers from zone data for which it has authority. A zone should have multiple authoritative servers with synchronized data and independent failure considerations. An authoritative-only server need not perform recursion for arbitrary clients.

CheckWhat makes a server authoritative for a zone?

It once queried the zone through a public resolver. It serves the zone data under the relevant delegation and configuration. It responds fastest to one ping.

Zones and Zone Storage

A zone is an administratively served portion of the DNS namespace. It begins at a zone apex and can delegate child zones. Zone data may be stored in a text zone file, generated from a database, loaded through an API, or synthesized by software; “zone file” is not a mandatory physical implementation.

The zone apex normally has an SOA record and an NS set. Delegation data at a parent identifies child authoritative servers, sometimes accompanied by glue address records needed to reach in-bailiwick server names.

CheckWhat is a DNS zone?

An administratively served portion of the namespace. A mandatory single text file on every client. An Ethernet broadcast domain identified by a VLAN.

Resource Record Fields

A resource record has an owner name, TTL, class, type, and type-specific RDATA. For example:

www.example.com.  300  IN  A  192.0.2.25

The owner is www.example.com., TTL is 300 seconds, class is Internet, type is IPv4 address, and RDATA is the address. Field omission and relative-name rules in zone-file syntax require careful origin handling.

CheckWhich record type publishes mail-exchanger preference and hostnames?

A NS MX

TTL and Negative Caching

Positive records use TTLs to limit cache reuse. Negative answers such as a proven nonexistent name can also be cached according to SOA-derived rules. Lowering a TTL shortly before a planned change affects only records fetched after caches observe the lower value; previously cached longer TTLs remain until expiry.

CheckWhy lower a DNS TTL well before a planned address change?

The TTL modifies the server's Ethernet MTU. A lower TTL guarantees the new application is healthy. Existing caches need time to expire records learned with the old longer TTL.

Summary

You can now separate DNS recursion, authority, namespace management, and cached records.

1. Identify stub and recursive resolver roles.
2. Define authority through delegated zone service.
3. Treat a zone as namespace responsibility, not one required file.
4. Read owner, TTL, class, type, and RDATA fields.
5. Plan cache lifetimes before DNS changes.