Skip to content
AdminKit

Stub resolver

What is Stub resolver?

A stub resolver is the lightweight DNS client built into an operating system that forwards every lookup to a recursive resolver rather than walking the DNS hierarchy itself, a pattern RFC 1123 explicitly describes. AdminKit's DNS resolver test walks you through checking which resolver a device's stub resolver is actually using.

By AdminKit team, Netbase JSC · Updated

A stub resolver is the lightweight DNS client built into an operating system or application that forwards every lookup straight to a recursive resolver, rather than walking the DNS hierarchy itself. ICANN’s own explainer puts it the same way: “A stub resolver is a very simple DNS client that relays an application’s request for DNS data to a more complicated component called a recursive resolver.” RFC 1123 (an Internet Standard per the IETF Datatracker) explicitly names this pattern, describing most end-host resolvers as minimal stub resolvers that depend on a nearby recursive resolver to do the iterative lookups RFC 1034 defines. Most operating systems ship a stub resolver pointed at whatever resolver address DHCP, a router, or an ISP supplies, though it can be repointed at a different one, including a Protective DNS or DNS-over-HTTPS resolver, to change what queries see and get filtered. On Linux, for example, the C library’s resolver reads the recursive resolvers to use from nameserver lines in /etc/resolv.conf, and the resolv.conf(5) manual page notes that up to three name servers may be listed. A stub resolver does no caching or validation on its own, so behavior like DNSSEC validation or content filtering depends entirely on the recursive resolver it’s configured to use. AdminKit’s DNS resolver test tool checks which resolver a device is actually using and how it answers. Because a stub resolver never walks the hierarchy itself, privacy techniques like QNAME minimization apply only at the recursive resolver it forwards to, not within the stub.

Stub resolver is part of AdminKit’s DNS cluster; the full AdminKit glossary lists other related terms.

Frequently asked questions

Does a stub resolver do its own caching or DNSSEC validation?

No -- a stub resolver does no caching or validation on its own, so behavior like DNSSEC validation or content filtering depends entirely on the recursive resolver it's configured to forward to.

Where does a stub resolver get the address of the resolver it forwards to?

Usually from DHCP, a router, or an ISP by default, though it can be repointed at a different one, including a Protective DNS or DNS-over-HTTPS resolver, to change what queries see and get filtered.

Is the stub resolver pattern formally defined anywhere?

Yes -- RFC 1123 explicitly names it, describing most end-host resolvers as minimal stub resolvers that depend on a nearby recursive resolver to do the iterative lookups RFC 1034 defines.

Does QNAME minimization apply inside the stub resolver itself?

No -- because a stub resolver never walks the DNS hierarchy itself, privacy techniques like QNAME minimization apply only at the recursive resolver it forwards to, not within the stub.