On 25/07/16 23:42, Alan Doherty wrote:
> its the sort of thing i think is not a case for acme having normal
> or short term certs but for acme to allow CA to provide fixed term 
> certs
> 
> provider/CA  X can decide themselves to profit from your
> use-scenario by offering normal user acme api (certs of nominal term)
> special user acme api (delivering all certs with a shorter fixed
> term)
> 
> rather than having a selector in the design of the protocol for user 
> custom terms [...]

Note that the current ACME draft already has "notBefore" and "notAfter"
fields for this purpose. The actual certificate lifetime a CA permits is
a policy decision and shouldn't be anything ACME dictates (which no one
is suggesting).

The question here is whether the kind of "Certificate Delegation" you'd
typically see with a CDN (or similar use-case) should be part of ACME,
or rather something that happens behind the scenes.

I'd lean towards Option 1. Regarding the pro arguments for Option 2:
1. It's not clear whether not submitting the actual certificates to CT
is feasible or even a good idea. How would logging only a delegation
ticket interact with something like Expect-CT (or CT stapling in
general)? Seems like this would require changes to various existing
software infrastructure in order to work. I also don't expect the 3-day
certificates to become the default, so this might be a bit of a
premature optimization with regards to CT submission load if only a
small number of high-profile targets decide to make use of it.
2. The domain owner-controlled "certificate management server" seems
like it would be a more flexible option in general. This would allow
organizations to add their own auditing and validation controls to the
process, as opposed to a simple "here's your delegation ticket, now go
talk to the ACME server." It also doesn't sound like a huge obstacle for
the kind of organization that would even consider the risk of CDN
delegation/revocation, IMO.

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

Reply via email to