Internet-Draft | MPIC | October 2024 |
Ahmad, et al. | Expires 24 April 2025 | [Page] |
This memo defines an API for Multi-Perspective Issuance Corroboration (MPIC) services to facilitate domain control validation (DCV) from multiple network perspectives. MPIC enhances the security of publicly-trusted certificate issuance by mitigating the risk of localized, equally-specific BGP hijacking attacks that can undermine traditional DCV methods permitted by the CA/Browser Forum Baseline Requirements for TLS Server Certificates. This API enables Certification Authorities (CAs) to more reliably integrate with external MPIC providers, promoting a more robust and resilient Web PKI ecosystem. The API design prioritizes flexibility, scalability, and interoperability, allowing for diverse implementations and deployment models. This standardization effort is driven by the need to consistently address vulnerabilities in the domain validation process highlighted by recent research and real-world attacks, as reflected in Ballot SC-067 V3 of the CA/Browser Forum's Server Certificate Working Group.¶
This note is to be removed before publishing as an RFC.¶
The latest revision of this draft can be found at https://open-mpic.github.io/draft-mpic/draft-westerbaan-secdispatch-mpic.html. Status information for this document may be found at https://datatracker.ietf.org/doc/draft-westerbaan-secdispatch-mpic/.¶
Discussion of this document takes place on the Security Dispatch Working Group mailing list (mailto:[email protected]), which is archived at https://mailarchive.ietf.org/arch/browse/secdispatch/. Subscribe at https://www.ietf.org/mailman/listinfo/secdispatch/.¶
Source for this draft and an issue tracker can be found at https://github.com/open-mpic/draft-mpic.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 24 April 2025.¶
Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
Before issuing a certificate to a subscriber, certificate authorities are required to validate that the subscriber indeed controls the domains on the certificate. For this purpose certificate authorities use various methods of domain control validation (DCV), including but not limited to via HTTP, DNS, and ALPN.¶
Several, but not all, CAs use the specific DCV methods of ACME [RFC8555]. Domain control validation is vulnerable to DNS and BGP hijacks. These can be partially mitigated by performing DCV from multiple network perspectives, which is dubbed "multiple perspective issuance corroboration" (MPIC). corroboration (MPIC) for domain control validation. Ballot SC-67 v3 of CA/B forum requires MPIC to be performed by all certification authorities (CAs) in the future.¶
Running MPIC requires maintaining a presence across the globe. For smaller CAs it may make sense to run a shared MPIC service or outsource it to a third party. This memo specifies a standardised API for such a usecase. Another usecase is for a CA to have a standby MPIC service in case its primary fails.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
The MPIC API implements a system for domain validation and CAA record checking using multiple perspectives across different regions.¶
An MPIC service can accept JSON requests over an arbitrary communication channel with an MPIC client. The MPIC service then produces MPIC responses which are sent back to the MPIC client over the communication channel. The communication channel may be synchronous (i.e., stall open on each MPIC request until an MPIC response is generated) or asynchronous (i.e., send a message to the MPIC service, close, and then receive a future message from the MPIC service containing a corresponding MPIC response). The MPIC service protocol does not provide any form of matching between MPIC requests and MPIC responses. A communication channel is responsible for ensuring that a client can match requests with corresponding responses. The communication channel MUST provide confidentiality and integrity as well as support for authentication of the MPIC service.¶
This document describes communication with an MPIC service using HTTPS POST as the communication channel. Alternate communication channels include gRPC, Apache Kafka, RabbitMQ, etc...¶
A MPIC service running over the HTTPS POST communication channel is identified by a HTTPS url.
As a running example, say https://mpc.example.com/staging
.¶
A client requests a MPIC validation from the service
by sending a POST request to the resource /mpic/draft-00
below the service URL.
In the running example https://mpc.example.com/staging/mpic/draft-00
.¶
[[ The final version of the API will use /mpic/v1
. Incompatible
versions of the draft will bump the -00
. ]]¶
The body of the HTTP POST is a JSON object that describs the MPIC request. The service will respond with a JSON object containing MPIC results.¶
There are three different MPIC validation methods, described below. The request
object has a method
field that allows to distinguish between each.¶
caa
validation method
A caa
requests asks the MPIC service to retrieve the relevant CAA DNS
records for a given domain from multiple perspectives.¶
This method has the following specific fields.¶
domain
(required, string): The domain to check the CAA records for.¶
An example request is given below.¶
POST /staging/mpic/draft-00 Host: mpc.example.com Content-Type: application/json { "method": "caa", "domain": "some.example.com" }¶
If successful (described in TODO REF below), the response object contains a
success
field set to true
, and an caa
field, which itself is an object
with two fields:¶
domain
The domain on which the CAA records were found. This could be
a parent domain of the requested domain
.¶
records
A list of base64 encoded CAA records.¶
An example of a response for a succesful validation.¶
{ "success": true, "caa": { "domain": "example.com", "records": ["AAVpc3N1ZWxldHNlbmNyeXB0Lm9yZw=="] }, "perspectives": { "jfk": {"success": true}, "fra": {"success": true}, "lis": {"success": true} } }¶
On failure, the response object will have the success
field set to false
,
and an error
field describing the error.¶
[[ TODO do we to define the possible errors, or at least assign some codes? ]]¶
An example of a response for an unsuccesful validation.¶
{ "success": false, "perspectives": { "jfk": {"success": true}, "fra": {"success": true}, "lis": {"success": false} }, "error": "LIS saw record 'xyz' on example.com which was not present from perspective LIS" }¶
[[ TODO How much information to return on error to help debug, and how structured should it be? I'd say it's good to be helpful, but it's bad to be structured as it's less readable. ]]¶
http-acme
method
http-acme
requests the MPIC server to perform ACME http-01 challenge validation
[RFC8555] for the domain's HTTP server from each distributed perspectives.¶
Performs a GET from multiple perspectives, and checks whether the body matches expectation. Optionally, it allows performing an additional CAA record lookup for the domain.¶
The request JSON object has the following specific fields.¶
domain_or_ip
(required, string): The domain name or IP address being verified.¶
token
(required, string): The token value defined in [RFC8555] Section 8.3.¶
key_authorization
(required, string): The Key Authorization defined in [RFC8555] Section 8.1.¶
caa_check
(optional, boolean): Performs CAA validation at the same time for the domain as described above. Defaults to true unless domain_or_ip
is an IP address.¶
POST /staging/mpic/draft-00 Host: mpc.example.com Content-Type: application/json { "method": "http-acme", "domain_or_ip": "some.example.com", "token": "base64_url_token", "key_authorization": "base64_url_token.base64url_Thumbprint_accountKey" "caa_check": false, }¶
The MPIC server constructs a URL by populating the URL template
[RFC6570], http://{domain_or_ip}/.well-known/acme-challenge/{token}
, and verifies that the resulting URL is
well-formed, before making a HTTP GET request to the URL from each vantage
point. Each perspective SHOULD follow redirects when dereferencing the URL.
The MPIC server verifies that the key_authorization
value provided by the client matches
with the body of the response received from each perspective.¶
If the above verifications succeeds, then the validation is successful. If the request fails, or the body does not pass these checks, then it has failed.¶
Along side, the MPIC server queries for the CAA records for the
domain_or_ip
if the caa_check
request parameter is set to "true".¶
If either HTTP or CAA validation (when requested) fail,
the response objects contains a top-level
success
field set to false
, and contains an error
field
that describes the error.¶
If both succeed, the response object contains a top-level success
field set to true
.¶
The response also contains an object (located under the key perspectives
) with keys that uniquiely identify the perspectives used in the reqest.
Each perspective is associated with an object that contains success
key pointing to a boolean value of true
or false
to indicate whether validation was successful at that perspective or not.¶
If a CAA check was requested, the response object will contain a top
level caa
field as described in Section 3.1.¶
An example of a response for a succesful validation with caa-check
set to
false.¶
{ "success": true, "perspectives": { "jfk": {"success": true}, "fra": {"success": true}, "lis": {"success": true} } }¶
An example of a response for a succesful validation with caa-check
set to
true.¶
{ "success": true, "perspectives": { "jfk": {"success": true}, "fra": {"success": true}, "lis": {"success": true} } "caa": { "domain": "example.com", "records": ["AAVpc3N1ZWxldHNlbmNyeXB0Lm9yZw=="] } }¶
An example of a response where the validation failed with caa_check set
to true, and the caa
method being successful.¶
{ "success": false, "perspectives": { "jfk": {"success": true}, "fra": {"success": true}, "lis": {"success": false} } "error": "HTTP found unexpected value at LIS perspective", }¶
Similarly example of a response where caa_check set to true, and both methods fail.¶
{ "success": false, "perspectives": { "jfk": {"success": false}, "fra": {"success": true}, "lis": {"success": false} } "caa": { "domain": "example.com", "records": ["AAVpc3N1ZWxldHNlbmNyeXB0Lm9yZw=="] } "error": "HTTP found unexpected value at LIS perspective. CAA check failed at the JFK perspective.", }¶
Describe usage of Authorization
header.¶
TODO Security¶
This document has no IANA actions.¶
TODO acknowledge.¶