[dns-operations] The (very) uneven distribution of DNS root servers on the Internet

[Patrick W. Gilmore]

On May 16, 2012, at 18:45 , Randy Bush wrote:

>>> One could logically assume that if a caching server is within a
>>> certain radius of a node geographically, they are likely able to
>>> route to it (country boundaries/geography may change this, but I did
>>> say roughly).
>> 
>> This assumption is what leads folks to do what Pingdom has done.  It
>> is a common mistake to assume network topology matches geo-political
>> topology.  In many case, regulatory regimes, business rules,
>> etc. result in the exact opposite.
> 
> my favorite is that some root server operators intentionally limit bgp
> announcement scope, so it does not even follow net topology, let alone
> geo.

BGP has no notion of performance, latency, etc.  In most cases, limiting announcements of a dense and widely distributed anycast deployment can greatly improve global performance.

Put another way: Applying some intelligence to the announcements is far superior than following "net topology".

-- 
TTFN,
patrick