> Those are two different things. dns-resolver discovery is to be used
> to locate a DNS resolver/server in last resort when no other information
> for finding this server is available.

Reading the document I don't see a separate step where the
DNS resolver/server is located. Instead I see that as a last
resort DNS queries are sent to a well-known site-local unicast
address.

> LLMNR (at least to my understanding) is to be used
> in last resort when the information had not been found in the DNS maps
> (meaning you already have found a DNS server/resolver).

As I understand the intent one goal is for LLMNR to work 
on a network consisting of just a few hosts, i.e. there might not
be any router or DNS resolver.
In that case a likely LLMNR usage end up being that 
when resolving host names as a last resort when DNS didn't return 
an answer, then use LLMNR.

Thus I can't tell which of these last resorts should be the last one.


> I thought that the IETF was in the business of specifying protocols,
> not how people should implement them nor how people should
> use them.
> The best we can do here is to providing guidelines like
> "use it only under those circumstances if you don't want to get hurt"

I think that specifying protocols and folks implementing them
is a bad idea unless the result is actually operationally useful.

Thus ignoring operational issues as you seem to advocate 
doesn't make for quality IETF standards.


--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to