Resolver behaviour with rrset

[Barry Margolin]

In article <c90o29$21u9$1 at sf1.isc.org>,
 "Sebastian Castro Avila" <secastro at nic.cl> wrote:

> On Wed, 26 May 2004 00:42:00 +0100, Jim Reid <jim at rfc1035.com> wrote:
> 
> >     Sebastian> How does it work if a server became unreachable?
> >
> > That server is considered to have an infinite round trip time. [For
> > some definition of "infinite".] So the dead server won't get queried
> > much after it became unreachable, all other things being equal. However
> > the resolving name server will periodically "forget" the dead server
> > has an infinite RTT and query it once more. Any response or
> > non-response is then used to update the server's idea of which name
> > server is closest for a given domain. This means a name server's
> > resolver is continually keeping track of the RTTs to the name servers
> > it queries and adjusting those RTTs in light of what is currently
> > going on.
> >
> 
> So, I should think that this resolver will choose a server based on a  
> "weighted list", where this weight is based on response time. The server  
> with better response time will be prefered, but it won't be the "only one"  
> used.

What I think happens is that every time a server is used, it is 
"penalized" a bit and its response time will be increased.  Eventually a 
good server will be penalized beyond some of the bad servers, so the 
latter will be tried again.

> I'm looking for some master thesis, and I'm exploring the posibility of  
> making a "model" or simulation to choose the right place and the right  
> capacity for a set of nameservers for a domain.

If you're a software engineering grad student, you should be able to 
study code to find out how it actually works.

-- 
Barry Margolin, barmar at alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***