Resolvers

Resolvers are software components that translate queries or identifiers into data or resources. In Internet infrastructure, DNS resolvers translate human-readable hostnames into IP addresses, enabling clients to locate services. DNS resolvers come in two primary modes: recursive and iterative. A recursive resolver accepts a query from a client and performs the full lookup chain on its own, querying root servers, top-level domain servers, and authoritative servers until an answer is obtained. An iterative resolver, by contrast, may return referrals to other servers, leaving part of the lookup to the client or forwarding agent.

During a lookup, the resolver first checks its local cache for a recent answer. If a cached entry is available and valid, it is returned immediately. If not, the resolver queries the DNS hierarchy: starting at a cache of root hints or root servers, then the appropriate TLD server, and finally the domain's authoritative servers. The final answer is cached with a time-to-live value to serve subsequent requests efficiently. DNSSEC adds cryptographic validation, allowing resolvers to verify that responses have not been tampered with.

Deployments vary: residential users often rely on ISP or public resolvers, while organizations may operate private resolvers for performance or policy reasons. Public resolvers frequently offer features such as caching, DoH (DNS over HTTPS) or DoT (DNS over TLS) to protect user privacy by encrypting queries between the client and the resolver. Privacy and security considerations include query logging, potential centralization, and the need to trust the resolver to enforce correct policies.

Beyond DNS, the term resolver also appears in other systems, such as dependency resolution in package managers or field resolvers in GraphQL, where a resolver maps a request to a computed value.