Reldnsprefetch

Reldns-prefetch, often written as reldnsprefetch, is a web performance optimization technique that proactively resolves DNS records for domains that are likely to be accessed from the current page, with emphasis on those related by a base or reference domain. The aim is to reduce the latency of future requests by warming the DNS cache ahead of user actions.

Mechanism: A client component analyzes the loaded document and its resources to identify candidate domains from hyperlinks, script sources, image URLs, and API endpoints. It derives a set of relative domains based on a reference domain—for example, subdomains of the current host or related base domains—and issues speculative DNS queries. Results are cached locally or at a resolver. The approach favors a small, policy-driven candidate set and may respect user preferences and organizational privacy rules.

Deployment: Implementations can appear in browsers, edge proxies, or CDN nodes as an opt-in or policy-driven feature. It often operates alongside traditional DNS prefetching and connection coalescing, extending prefetching to relative domain relationships and potentially using DNS records such as A, AAAA, and CNAMEs. Security and privacy considerations—such as potential exposure of visit patterns and increased query load—are addressed via limits, rate controls, and user-configurable settings.

Limitations and status: As of current discourse, reldnsprefetch is described in performance literature and vendor white papers rather than as an official standard. Adoption varies and depends on integration with existing DNS privacy and transport security practices, such as DNS over HTTPS or DNS over TLS.