In the early days of the Internet, computers were identified solely by numerical addresses. While machines handled these identifiers easily, humans did not. As networks expanded and the number of connected systems grew, relying on static host files and memorised numbers quickly became unmanageable. The Internet needed a scalable naming system that could grow with it.
Paul Mockapetris solved this problem by designing the Domain Name System. DNS transformed the Internet into something people could actually use. It introduced a hierarchical, distributed naming architecture that still underpins every web request, email delivery, and network service today.
Life before DNS
Before DNS, the ARPANET relied on a centrally maintained text file called HOSTS.TXT. This file mapped hostnames to numerical addresses and was distributed manually to connected systems. As more machines joined the network, updating and synchronising this file became a logistical problem. Conflicts, outdated entries, and delays were common.
The centralised model did not scale. It also created a single point of failure. The Internet needed a decentralised naming solution that matched its distributed design philosophy.
Designing a distributed naming system
Mockapetris approached the problem in 1983 while working at the University of Southern California’s Information Sciences Institute. Instead of one global list, he proposed a hierarchical namespace that could be delegated across organisations and regions.
DNS introduced several key ideas:
- A tree-structured namespace with domains and subdomains
- Distributed authority, where each domain could manage its own records
- Caching, to reduce query load and improve performance
- Redundancy, through multiple authoritative servers
This design allowed the system to grow organically, without requiring central coordination for every update.
How DNS works
When a user enters a domain name, their system does not contact a single central server. Instead, it queries a series of servers, starting from root servers, moving through top-level domains, and finally reaching the authoritative server for the requested name.
This layered approach ensures resilience. If one server is unavailable, others can respond. Cached responses further reduce load and improve reliability.
Mockapetris documented DNS in a set of RFCs that became foundational references. These specifications remain relevant, with only incremental updates and extensions added over time.
Why DNS was a turning point
DNS made the Internet accessible beyond technical users. It enabled:
- Memorable domain names instead of numerical addresses
- Independent administration by organisations and service providers
- Rapid expansion of services without central bottlenecks
- A stable platform for future protocols and applications
A system under constant pressure
Although DNS was designed for openness, it has also become a target for abuse and attack. Spoofing, cache poisoning, and denial-of-service attacks have all exploited its critical role. In response, extensions such as DNSSEC were developed to add cryptographic verification.
These adaptations show the strength of Mockapetris’ original design. DNS could be extended and reinforced without being replaced.
DNS operates in the background, often unnoticed until something goes wrong. Yet it is one of the most critical components of the Internet’s infrastructure. Every website visit, every email exchange, and every API call depends on it.