[dns-operations] Resolvers choosing low latency nameservers

[Matthew Pounsett]

On 2013/06/21, at 09:27, Matthäus Wander wrote:

> Hi,
> 
> are there any studies or anecdotal evidence about how recursive
> resolvers select a query destination from a set of authoritative servers
> with known RTTs, and how often they re-probe the slower ones?
> 
> Specifically, how many queries in what period of time would it take
> until a BIND or Unbound has re-probed all auth ns of e.g. com?

This may not answer your second question, but at the DNS-OARC meeting in London last spring there was a presentation covering your first.

<https://www.dns-oarc.net/files/workshop-201203/OARC-workshop-London-2012-NS-selection.pdf>