Hi Corey,

thank you for the comments.

So, we have an allowed solution, which provides three different certificate
status answers based on the verifier's software:
based on the CARL:
- good
- revoked
- crashed.

And it fulfills the BRG, especially these:
4.10.1 Operational characteristics
Revocation entries on a CRL or OCSP Response MUST NOT be removed until
after the Expiry Date of the revoked Certificate.

4.9.7 CRL issuance frequency
As issuing CA Certificates:
1. MUST update and publish a new CRL at least every twelve (12) months;
2. MUST update and publish a new CRL within twenty-four (24) hours after
recording a Certificate as revoked.

CAs MUST continue issuing CRLs until one of the following is true:
- all Subordinate CA Certificates containing the same Subject Public Key
are expired or revoked; OR
- the corresponding Subordinate CA Private Key is destroyed.

And ETSI 319 411-1:  Where CARL is used:
• CSS-6.3.9-12 [CONDITIONAL]: If CARL is used, a new CARL shall be
generated at least once a year with a nextUpdate of at most 1 year after
the issuing date.

By the way, which nextUpdate should be inserted into the CARL containing
revoked Root CA certificate if
- not all subordinate CA certificates are revoked, or
- all subordinate CA certificates are revoked?

And I think one new paragraph would be very useful in the next BRG about
the revocation of the Root CA certificates, if it is a really allowed
method.

Thank you in advance!

Best Regards,
Peter



On Wed, Feb 14, 2024 at 11:40 PM Corey Bonnell <[email protected]>
wrote:

> Hi Peter,
>
> Comments inline.
>
>
>
> > What do you think, is it an acceptable practice, if the Root CA revokes
> its self-signed certificate and puts this revocation information into the
> CARL signed by the (revoked(?)) Root CA private key? (Subscribers of the
> NR-CA are the subordinate CA owners only.)
>
>
>
> I think it’s acceptable and currently allowed, but largely ceremonial and
> insufficient for Relying Parties (clients) to distrust the root
> certificate. Checking the status of self-signed root certificates used as
> trust anchors is largely never done by clients. Additionally, revoking the
> certificate does not relieve the CA’s obligation to continue operating the
> root CA in accordance with browser policy, including periodic audits. The
> CA would have to request removal of the CA from the browser trust store and
> wait for its removal to be relieved of the ongoing compliance requirements.
>
>
>
>    - How to use this "suicide CRL"?
>
>
>
> Peter answered this quite well, although I have heard reasonable arguments
> that the root CA certificate appearing on a CRL that the CA itself issued
> is the clearest signal that it is revoked. It’s on the CRL either due to
> explicit action by the CA, or an attacker has gained control of the root
> key material and is using the key material to issue a CRL that revokes the
> certificate. In either case, the root should no longer be trusted.
>
>
>
> Thanks,
>
> Corey
>
>
>
> *From:* [email protected] <[email protected]> *On
> Behalf Of *Peter Mate Erdosi
> *Sent:* Tuesday, February 13, 2024 4:59 AM
> *To:* [email protected]
> *Cc:* Tavis Ormandy <[email protected]>; Matthew Hardeman <
> [email protected]>; [email protected] <
> [email protected]>; Jeremy Rowley <
> [email protected]>
> *Subject:* Re: BR revocation question
>
>
>
> Dear all,
>
>
>
> I have a question about  the revocation of the root certificate. I have
> not found "Reasons for Revoking a Root CA Certificate" chapter in the BRs.
>
>
>
> I read the following in this CPS (
> https://ecac.pki.gov.pk/repository/cps/ECAC_Certification_Authorities_CP_CPS_v1.4.pdf
> )
>
>
>
> "1.4.1 Appropriate Certificate Uses
>
> For certificate issued to the NR-CA itself: it is a special class of
> self-signed certificate that being the trust anchor of the Pakistan PKI.
> The NR-CA certificate can be used for (...):
> - Sign CRLs containing the list of subscribers’ revoked certificates and
> of NR-CA revoked self-signed certificates,"
>
> where "NR-CA" means "National Root CA".
>
>
>
> What do you think, is it an acceptable practice, if the Root CA revokes
> its self-signed certificate and puts this revocation information into the
> CARL signed by the (revoked(?)) Root CA private key? (Subscribers of the
> NR-CA are the subordinate CA owners only.) How to use this "suicide CRL"?
>
>
>
> I read in several places that the root CA certificate cannot be revoked.
> (e.g.
> https://security.stackexchange.com/questions/90254/can-a-rootca-be-revoked
> )
>
>
>
> Thank you in advance for any comments!
>
>
>
> Best Regards,
>
> Peter
>
>
>
> Tavis Ormandy a következőt írta (2022. augusztus 10., szerda, 21:10:13
> UTC+2):
>
> On Wed, Aug 10, 2022 at 06:57:32PM +0000, Jeremy Rowley wrote:
> > The general instruction I got was you couldn’t use revocation as a
> threat to keep customers from switching CAs. That was pretty clear from
> Ryan. Other bad actions were implied as prohibited, like revocation just
> because a contract terminated. Like I said, I’d love to see it written down
> as an official policy as the bounds and applicability are hearsay.
>
> Thanks, I'll send some emails!
>
> On Wed, Aug 10, 2022 Matthew Hardeman wrote:
> > This sounds less like it was about a customer amidst migration and more
> like it was a "sell long validity cert on `credit` and collect payment over
> cert lifetime".
>
> Hardly. Regardless, is revocation intended to protect trust in the
> ecosystem, or as an asset recovery and reposession tool for the CA
> industry?
>
> I think it's the former.
>
> Tavis.
>
> --
> _o) $ lynx lock.cmpxchg8b.com
> /\\ _o) _o) $ finger [email protected]
> _\_V _( ) _( ) @taviso
>
> --
> You received this message because you are subscribed to the Google Groups "
> [email protected]" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/4607eddf-5d61-4bdb-b589-fc19796d7f03n%40mozilla.org
> <https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/4607eddf-5d61-4bdb-b589-fc19796d7f03n%40mozilla.org?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"[email protected]" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/CADuWVBXXoPqiGSd2z%3DFdYMR9YUBFKa5%3D4%2BwGWh4%2B6nwMNxW0KA%40mail.gmail.com.

Reply via email to