On Jun 24, 2013, at 10:15 AM, Joe Abley <[email protected]> wrote:

> 
> On 2013-06-24, at 12:56, Paul Hoffman <[email protected]> wrote:
> 
>> In looking at the diffs, I still have the question I had earlier about NSID. 
>> I give the following command three times quickly:
>>  dig -4 @L.ROOT-SERVERS.NET . SOA +nsid +norec +noall +comments
>> I get three different answers (lax16..., lax09..., lax13...).
>> 
>> This is similar to the variance I see when doing
>>  dig -4 @L.ROOT-SERVERS.NET HOSTNAME.BIND CH TXT +short
>> and
>>  dig -4 @L.ROOT-SERVERS.NET ID.SERVER CH TXT +short
>> However:
>>  dig IDENTITY.L.ROOT-SERVERS.ORG TXT +short
>> gives me the same answer each time I give it.
>> 
>> So, I'm still confused about the paragraph just before section 4.1.
> 
> That paragraph addresses the case of:
> 
> (a) send query, get problematic response
> (b) do HOSTNAME.BIND (or similar) query to identify what node I'm talking to
> 
> It's entirely possible that the responding node for (a) will be different 
> from the responding name for (b), in which case the node identity wouldn't 
> correspond with the node that seemed to be giving a problematic response.
> 
> If instead you do:
> 
>  (a) send query with NSID, get problematic response
> 
> then the identity of the node exposed by the use of NSID corresponds exactly 
> to the node that supplied the problematic response, in all cases, guaranteed.

Ah, got it. I missed the linkage of "add NSID to the troublesome query". 
Proposal:
Current:
Use of the NSID option can obviate this possibility
Proposed:
Adding the NSID option to the query that causes the problematic response can 
obviate this possibility
_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to