Re: [Inip-discuss] Innovation in DNS

[Andrew Sullivan <[e-mail]>]

On Thu, Aug 25, 2016 at 05:01:46PM +0100, Bauyrzhan Askar wrote:
> For the purpose of clarification of the claim, TTL specifies time it will
> be in cache of DNS resolver. I  point you to “request to *unknown domain
> name*” and it means that there is no record data for the unknown domain
> name not just on Internet users PC but also in DNS resolvers cache.

Perhaps I'm misunderstanding you, but it sounds like you think that if
unknown.example.com arrives ar a resolver, it needs to do the whole
resolution chain.  But the fact is that the NS set for com and
possibly for example.com will both be in the cache.  There's no need
to ask the root about com in this case.  Com is almost always hot in
every cache.  This is true for a lot of TLDs.  And with the advent of
the signed root, lots of resolvers just copy the root, so they don't
really have to consult the root servers at all.

> As for the risk it brings, what kind of risk are you mentioning?
> Cybersquatting, risk of switching to AINA System or other risks?

You're changing the system, and making it considerably more complex.
That necessarily brings risk.

> I want to draw your attention to the part where implementation of the AINA
> System is described. As you can see for the purpose of simplicity,
> switching to AINA System can be conducted through three scenarios. (at
> least three and it is up to imagination). The first one is the one which
> has no risk, and what needed is just to direct requests to AINA System in
> case of unknown domain name for root nameservers.

There's not "no risk" to that.  You're introducing a new set of
dependencies.  That automatically brings risk, and I don't see what
the reward is.

Best regards,

A

-- 
Andrew Sullivan
[e-mail]