Re: [DNSOP] KSK rollover choices

[Joe Abley <[e-mail]>]

Hi Paul,

On 31 Oct 2018, at 14:54, Paul Vixie <[e-mail]> wrote:

> Jim Reid wrote:
> 
>>> On 31 Oct 2018, at 00:27, Mark Andrews<[e-mail]>  wrote:
>>> 
>>> Bootstrap is still a issue.  Over fast TA rolling makes it more of
>>> a issue.
>> 
>> Indeed. And that's the underlying problem that needs to be fixed IMO
>> - for instance when/if there's an emergency rollover.
> 
> bootstrappers should have https access to a complete history of root ksk, each one signed by its predecessor. this doesn't handle revocation, but nothing in dnssec handles revocation, and that's by design, and so i'm inclined not to worry about it.

The existing scheme provides bootstrappers with a complete history of trust anchors (published by digest). The published object was intended to be accompanied by a detached signature that could be trusted by a vendor, since there was to be a certificate chain from the object-signing certificate back to an X.509 trust anchor that made sense to the client (e.g. an iOS code-signing key, a Windows code-signing key, an ISC code-signing key, etc, etc). This is described in RFC 7958.

The only "vendor" that ever published such a certificate was the (largely proof-of-concept) "ICANN" vendor.

I'm not suggesting that the scheme described in that document is good, or that it shouldn't be replaced with something better, but it's one possible starting point for the discussion.


Joe