CSC 2231 - DNS review

Development of the Domain Name System
--------------------------------------
Paul Mockapetris, Kevin Dunlap

This paper describes the evolution of DNS after the system was
initially designed, in 1983. The paper addresses issues like the
need for the system, design, implementation status at that time,
surprises, successes and shortcomings of the system.

The DNS came to replace a centralized system (HOSTS.TXT) and its
success mainly relied on the need for a distributed naming system
in a growing distributed environment (Internet). DNS is hierarhichal
and has two major components: name servers (repositories of
information) and resolvers (interface with the clients). The tree name
space is "divided" into zones controlled by a specific organization.
The paper also emphasizes the use of caching, both positive as well
as negative.

As a paper weakness, the authors didn't take into consideration
security issues in their system design. But considering that the
paper is from 1988, it's understandable.

Taking into account the fact that DNS is the naming system being
used now I have to say that, although the paper is old, it represents
a good practical paper. The authors consider both strengths and
weaknesses of the system, giving a fair description of it. They also
leave space for future improvements. The first paragraph in the
Conclusions shows this "open to debate" aspect of the paper.
Received on Mon Oct 03 2005 - 10:35:34 EDT