Stub resolver
The small DNS client built into an operating system or application.
The Internet’s naming system
DNS is not merely a phone book. It is a distributed database, a delegation system, a cache hierarchy, a security boundary and—more often than people realise—part of an application’s control plane.
A system of distinct roles
A client’s stub resolver asks a recursive resolver for a final answer. The recursive resolver consults the hierarchy, caches what it learns and may validate DNSSEC. Authoritative servers publish the data for zones; they do not normally browse the hierarchy on a client’s behalf.
The small DNS client built into an operating system or application.
Finds the answer, caches it and optionally applies validation or policy.
Publishes signed or unsigned zone data and referrals.
Explore the guide
Each page focuses on one layer, while cross-links show where the layers meet.
Follow an iterative lookup through the root, TLD and authoritative layers; understand TTLs, referrals and negative caching.
Trace a lookup →Learn how DS, DNSKEY, RRSIG and NSEC records build a chain of trust—and what DNSSEC does not encrypt.
Validate the chain →Compare classic DNS, DoT, DoH and DoQ, then place QNAME minimisation and encrypted transports in the correct threat model.
Compare transports →Read zone-file syntax and understand the operational meaning of A, AAAA, CNAME, MX, NS, SOA, TXT, CAA, TLSA and SVCB/HTTPS.
Browse record types →Separate recursive, authoritative, filtering, encrypted-transport and signing responsibilities into clear trust boundaries.
Design the stack →Use concrete checks, safe deployment sequences, secondary servers, TSIG, zone transfers and monitoring to keep DNS dependable.
Operate it well →Build a practical authoritative NSD service, then confine it with AppArmor, systemd sandboxing, narrow capabilities and controlled writable paths.
Harden an authority →Three separate questions
DNSSEC lets a validating resolver detect forged or modified signed data.
DoT, DoH and DoQ encrypt a particular client-to-resolver hop.
Redundant authority, sane TTLs, monitoring and disciplined changes support availability.