[dns-operations] Validation direction (Was: Re: Org Dnskey TTL)

[Joe Abley]

On 2010-04-20, at 10:43, Edward Lewis wrote:

> At 10:14 -0400 4/20/10, Joe Abley wrote:
> 
>> It had not actually occurred to me that anybody would think bottom-up was
>> a good idea.
> 
> Gulp!  That was the intention of the protocol design!!!

Perhaps context is necessary -- I'm talking about the impact on query patterns from validation, not any internal algorithm a validator might use to validate an answer. The thread began with pondering about the practical impact of low-TTL DNSKEY RRSets in ORG and wandered off into a corner to muse whether the impact would vary with the number of secure delegations from that zone.

In practical terms, given that in reality every validator sets DO=1 for every query and that resolution is always top-down following the initial priming query, it seems unnatural to think that query order would be anything other than top-down.

And to define my terms and make sure we're all holding the map the same way, as viewed by me the root is up and www.deep.down.in.the.namespace.away.from.the.root is down.


Joe