[dns-operations] Test on Priming Behavior

[Paul Hoffman]

On Dec 23, 2014, at 2:43 AM, Tony Finch <dot at dotat.at> wrote:
> 
> Davey Song <songlinjian at gmail.com> wrote:
>> 
>> But  I do not find any specification on the priming process of resolver,
> 
> There is a draft
> http://tools.ietf.org/html/draft-ietf-dnsop-resolver-priming

That long-expired draft says:

   A priming query SHOULD use a QNAME of "." and a QTYPE of NS.

Davey's question appears to be about implementations that don't seem to do that, but instead just send a first query with RD=1 to a server on configured priming list.

Mark Andrews' response was "If you are going to trust them to give you the root NS RRset you can trust them to give you a TLD referral", which seems about right to me. That is, is there really a reason for starting the cache with a query for ". IN NS" instead of just "whatever IN A"?

--Paul Hoffman