named

[richard.pieri-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org (Richard Pieri)]

On Dec 5, 2010, at 10:03 PM, Tom Metro wrote:
> 
> I don't think that's correct. It subverts the intent of DNS being a
> distributed database.

A distributed database is a single database with storage replicated across many discrete storage systems.  GIT is a distributed database.  A GIT replica contains the entire set of data (source code).  DNS is not a distributed database.  DNS is an hierarchical naming system built on top of many, many distributed databases.  Phrased another way, BIND is a distributed database engine and DNS is an application built on top of that engine.


> If it were true that it was hard to find a public recursive server that
> was fast, reliable, and didn't monkey with the records, then running
> your own recursive resolver would make sense. But that's not the case.

And is immune to cache poisoning.  Most public servers -- particularly ISPs and Google -- run caching name servers.  Authoritative servers like the root name servers are usually non-caching, thus cannot be poisoned, thus ensuring that you always get the correct records for your queries.

--Rich P.