Inline below.

- JL

On 4/10/12 10:06 AM, "Shane Kerr" <[email protected]> wrote:

>Jason,
>
>On Tuesday, 2012-03-27 15:19:23 +0000,
>"Livingood, Jason" <[email protected]> wrote:
>> I posted a ­01 a short time ago
>> (http://tools.ietf.org/html/draft-livingood-negative-trust-anchors-01)
>> and listed this in open items for the next version. I may have some
>> questions on the best way to include that.
>
>I like this idea; it was something I planned on including in BIND 10
>for a while, although we've got other higher-priority things to add
>first (like DNSSEC validation). ;)
>
>One tiny bit:
>
>   Furthermore, a Negative Trust Anchor should be used only for a short
>   duration, perhaps for a day or less.
>
>
>While for larger ISPs a top-100 approach makes sense, and so a 1 day fix
>is reasonable to expect, other operators will have other concerns. For
>example, you may have a distributor for your business that has a crappy
>IT department and simply can't get their zones fixed in a reasonable
>time, or you may have a department at your university that is literally
>gone for the summer.

This is indeed an issue we have observed. Here are some recent examples
where we've not seen a 1 or 2 day resolution:

http://dnsviz.net/d/www.bayfieldelectric.com/T4wofg/dnssec/


http://dnsviz.net/d/www.bayfieldelectric.com/T4wofg/dnssec/


>The approach I had planned on taking is simply to require that an
>administrator specify the ending time of the Negative Trust Anchor. If
>they want to, of course they can put 30 years (or perhaps however much
>time is left until their retirement), but at least they would have had
>to think about the issue!

I think that is a great approach! And perhaps have a default TTL in there
for the NTA of X hours.

- Jason






_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to