Internet-Draft TLS Trust Expressions October 2023
Benjamin, et al. Expires 21 April 2024 [Page]
Workgroup:
Transport Layer Security
Internet-Draft:
draft-davidben-tls-trust-expr-00
Published:
Intended Status:
Standards Track
Expires:
Authors:
D. Benjamin
Google LLC
D. O'Brien
Google LLC
B. Beck
Google LLC

TLS Trust Expressions

Abstract

This document defines TLS trust expressions, a mechanism for relying parties to succinctly convey trusted certification authorities to subscribers by referencing named and versioned trust stores. It also defines supporting mechanisms for subscribers to evaluate these trust expressions, and select one of several available certification paths to present. This enables a multi-certificate deployment model, for a more agile and flexible PKI that can better meet security requirements.

About This Document

This note is to be removed before publishing as an RFC.

The latest revision of this draft can be found at https://davidben.github.io/tls-trust-expressions/draft-davidben-tls-trust-expr.html. Status information for this document may be found at https://datatracker.ietf.org/doc/draft-davidben-tls-trust-expr/.

Discussion of this document takes place on the Transport Layer Security Working Group mailing list (mailto:tls@ietf.org), which is archived at https://mailarchive.ietf.org/arch/browse/tls/. Subscribe at https://www.ietf.org/mailman/listinfo/tls/.

Source for this draft and an issue tracker can be found at https://github.com/davidben/tls-trust-expressions.

Status of This Memo

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 21 April 2024.

Table of Contents

1. Introduction

TLS [RFC8446] endpoints typically authenticate using X.509 certificates [RFC5280]. These certificates are issued by a certification authority (CA) and associate the TLS public key with some application identifier, such as a DNS name. If the CA is among those trusted by the peer, it will accept this association. The authenticating party is known as the subscriber and the peer is the relying party.

Subscribers typically provision a single certificate, which must meet all requirements of all supported relying parties because relying parties do not communicate which CAs are trusted. This imposes significant costs on CAs, subscribers, and relying parties:

Deploying multiple certification paths would avoid these issues. However, relying parties and subscribers must then negotiate which to send in each connection. Section 4.2.4 of [RFC8446] defines the certificate_authorities extension, but it is often impractical. Some trust stores are large, and the X.509 Name structure is inefficient. For example, as of August 2023, the Mozilla CA Certificate Program [MOZILLA-ROOTS] would encode 144 names totaling 14,457 bytes.

This document defines a TLS extension and supporting mechanisms that allow relying parties and subscribers to succinctly communicate supported trust anchors to subscribers, using pre-shared information provisioned by root programs and CAs, respectively. This enables a multi-certificate deployment model, which supports a more flexible, robust Public Key Infrastructure (PKI). Section 9 discusses several deployment use cases that are directly improved by this model.

2. Conventions and Definitions

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.

This document additionally uses the TLS presentation language defined in Section 3 of [RFC8446].

2.1. Time

Times in this document are represented by POSIX timestamps, which are integers containing a number of seconds since the Epoch, defined in Section 4.16 of [POSIX]. That is, the number of seconds after 1970-01-01 00:00:00 UTC, excluding leap seconds. A UTC time is converted to a POSIX timestamp as described in [POSIX].

Durations of time are integers, representing a number of seconds not including leap seconds. They can be added to POSIX timestamps to produce other POSIX timestamps.

The current time is a POSIX timestamp determined by converting the current UTC time to seconds since the Epoch as per Section 4.16 of [POSIX]. One POSIX timestamp is said to be before (respectively, after) another POSIX timestamp if it is less than (respectively, greater than) the other value.

2.2. Terminology and Roles

This document discusses four roles:

Subscriber:

The party whose identity is being validated in the protocol. In TLS, this is the side sending the Certificate and CertificateVerify message.

Relying party:

The party that authenticates the subscriber. In TLS, this is the side that validates a Certificate and CertificateVerify message.

Certification authority (CA):

The service that issues certificates to the subscriber.

Root program:

An entity that manages a set of common trusted CAs for a set of relying parties.

While the first three roles are common to most X.509 deployments, this document discusses the role of a root program as distinct from the relying party. Trust decisions are often common between related relying parties, such as multiple clients running the same application. The root program is the entity making those common decisions, such as the software vendor for the application or operating system.

Additionally, there are several terms used throughout this document to describe this proposal:

Trust anchor:

A pre-distributed public key or certificate that relying parties use to determine whether a certification path is trusted.

Trust store:

A collection of trust anchors managed by the root program.

Certification path:

An ordered list of X.509 certificates starting the target certificate. Each certificate is issued by the next certificate, except for the last, which is issued by a trust anchor.

Trust store manifest:

A structure describing a series of versioned trust stores.

Trust expression:

A compact description of a relying party's trust store contents, referencing named, versioned trust stores.

CertificatePropertyList:

A structure associated with a certification path, containing additional information from the CA, for use by the subscriber when presenting the certification path.

TrustStoreInclusionList:

A property in the CertificatePropertyList which describes the corresponding trust anchor's inclusion in trust stores. This is used by the subscriber when evaluating trust expressions.

3. Overview

In the TLS handshake, a client (respectively, server) relying party sends trust expressions, which reference named, versioned trust stores to describe a list of trust anchors. The subscriber uses this information to select a certification path to return.

These structures are intended to be provisioned by root programs and CAs as follows:

By providing accurate trust anchor negotiation, this process avoids the need for relying parties to perform path-building [RFC4158]. Certification paths can be pre-constructed to end at one of the relying party's supported anchors.

4. Trust Store Manifests

This section defines a trust store manifest, which is a structure published by the root program to some well-known location. These manifests are used to compute certificate properties for the subscriber (Section 5) and trust expressions for the relying party (Section 6).

Trust store manifests are identified by a short, unique trust store name and define a list of trust store versions. A trust store version contains a set of trust anchors, and is identified by a name (from the manifest) and a version number.

[[TODO: Trust store names need to be unique, or at least unique among relying parties and subscribers that talk to each other, but also short. How do we allocate them? Registry? Just pick values? OIDs?]]

A trust store manifest is a JSON object [RFC8259] containing:

name:

A string containing a short, unique name that identifies the collection of trust stores

max_age:

A non-negative integer containing the number of seconds that this document may be cached.

trust_anchors:

A non-empty object describing a set of trust anchors. Each member's name is a short identifier for the trust anchor, used in entries below. Each member's value is an object describing the trust anchor, containing:

type:

A string describing the type of trust anchor.

In this document, the value of type is always "x509". In this case, the object additionally contains a member data whose value is a string containing a base64-encoded [RFC4648], DER-encoded [X690] X.509 certificate.

Future documents may extend this format with other trust anchor types. Recipients MUST ignore trust anchors with unrecognized type.

versions:

A non-empty array describing versions of the trust store. Each element of the array is a JSON object containing:

timestamp:

An integer, which is the time at which this trust store version was defined, as a POSIX timestamp (see Section 2.1).

entries:

A non-empty array describing the contents of this version of the trust store. These are known as trust store entries. Each element is an object containing:

id:

A string containing the name of some member of the trust_anchors object.

labels:

A non-empty array of non-negative, 24-bit integer labels associated with the trust anchor. See Section 4.2 for how this field is interpreted.

max_lifetime:

A non-negative integer containing the maximum lifetime, in seconds, of certification paths that this trust anchor is expected to issue.

Versions in versions are identified by their zero-indexed position in the array. That is, the first entry is version 0, the second is version 1, and so on.

Recipients MUST ignore JSON object members with unrecognized names in each of the objects defined above. Future documents MAY extend these formats with new names. Section 13 contains a CDDL [RFC8610] describing the above structure.

When updating a trust store manifest, root programs append a new object to the versions array to represent the new trust store version. When the new object references a trust anchor, the root program uses the existing members of trust_anchors, or adds new members as required. Updates MUST NOT modify or remove existing entries in the versions array.

4.1. Trust Store Entry Expiration

The max_age and max_lifetime fields define an expiration time for trust store entries in versions other than the latest version. For a trust store entry in version V, the expiration time is the sum of:

  • The timestamp of the subsequent version, i.e. version V+1

  • The max_age of the manifest

  • The max_lifetime of the trust store entry

Expiration times for entries in the latest version are not defined. They are determined once the root store appends a new version.

Trust store entries are not removed from their containing version after they expire. Rather, the expiration time is the point at which all unexpired certificates have incorporated information about subsequent trust store versions, per Section 5.2. This ensures the procedures in Section 6.3 and Section 6.5 interoperate.

4.2. Labels

Trust store entries reference labels, which are 24-bit integers that identify subsets of a trust store version. These integers are defined relative to the trust store manifest and are not globally unique. Root programs allocate labels when defining versions of a trust store.

Labels are used in trust expressions (Section 6.1) to exclude trust anchor entries from the expression. A label may identify an individual trust anchor if it is the only one with that label, or multiple trust anchors if they share the label.

The root program's label allocation scheme determines which sets may be efficiently represented. In particular, when trust anchors are removed in a later version of a trust store, trust expressions must express the removed set with labels as described in Section 6.5. Root programs SHOULD allocate individual labels for each trust anchor, and shared labels for trust anchors that will likely be removed together. For example, the root program may allocate shared labels for trust anchors with the same operator, or trust anchors that use some signature algorithm if it expects some relying parties to exclude by algorithm.

When allocating labels, root programs MAY repurpose a previously allocated label value after all previous entries referencing it have expired (Section 4.1).

5. Certificate Properties

In order to evaluate references to trust stores, e.g. in Section 6.1, subscribers require information about which trust stores contain each candidate certification path's trust anchor. This document introduces an extensible CertificatePropertyList structure to carry this information.

CertificatePropertyLists are constructed by CAs and sent to subscribers, alongside the certification path itself. They contain a list of properties the subscriber may use when presenting the certificate, e.g. as an input to certificate negotiation (Section 6.4).

A CertificatePropertyList is defined using the TLS presentation language (Section 3 of [RFC8446]) below:

enum { trust_stores(0), (2^16-1) } CertificatePropertyType;

struct {
    CertificatePropertyType type;
    opaque data<0..2^16-1>;
} CertificateProperty;

CertificateProperty CertificatePropertyList<0..2^16-1>;

The entries in a CertificatePropertyList MUST be sorted numerically by type and MUST NOT contain values with a duplicate type. Inputs that do not satisfy these invariants are syntax errors and MUST be rejected by parsers.

This document defines a single property, trust_stores, which describes trust store inclusion. Future documents may define other properties for use with other mechanisms.

Subscribers MUST ignore unrecognized properties with CertificatePropertyType values. If ignoring a property would cause such subscribers to misinterpret the structure, the document defining the CertificatePropertyType SHOULD include a mechanism for CAs to negotiate the new property with the subscriber in certificate management protocols such as ACME [RFC8555]. For example, Section 7.1 defines a "trustExpressions" ACME order field.

5.1. Trust Store Inclusion

The trust_stores property specifies which trust stores contain the certification path's trust anchor. Its data field contains a TrustStoreInclusionList, defined below:

enum {
    previous_version(0),
    latest_version_at_issuance(1)
} TrustStoreStatus;

struct {
    opaque name<1..2^8-1>;
    uint24 version;
} TrustStore;

struct {
    TrustStore trust_store;
    TrustStoreStatus status;
    uint24 labels<1..2^16-1>;
} TrustStoreInclusion;

TrustStoreInclusion TrustStoreInclusionList<1..2^16-1>;

Each TrustStoreInclusion describes a trust store version which contains this certification path's trust anchor. trust_store specifies a version of the trust store, and labels specifies the trust anchor's labels within that trust store.

If status is latest_version_at_issuance, trust_store was the latest trust store of its name at the time the certificate was issued. In this case, the trust expression evaluation algorithm (Section 6.3) predicts this information additionally applies to all versions greater than trust_store's version, up to the expiration of the certification path.

A TrustStoreInclusionList MUST satisfy the following invariants:

  • Each TrustStoreInclusion has a unique trust_store value.

  • The TrustStoreInclusion structures are sorted, first by length of trust_store's name, then by trust_store's name lexicographically, then by trust_store's version.

  • If status is latest_version_at_issuance in some TrustStoreInclusion, no trust_store with the same name but higher version appears in the list.

5.2. Computing Trust Store Inclusions

Each CA regularly fetches trust store manifests from root programs in which it participates, caching the most recently fetched manifest from each root program. When issuing a certification path to a subscriber, it runs the following procedure on each root program's trust store manifest, to compute a list of TrustStoreInclusions:

  1. If the cached manifest was fetched more than max_age seconds ago, fetch and cache a new copy.

  2. For each trust store version in the cached manifest's versions:

    1. Look for a trust store entry whose id, when indexed into trust_anchors, matches the certification path's trust anchor.

    2. If found, output a TrustStoreInclusion structure:

      • Set trust_store to the manifest's name, encoded in UTF-8 [RFC3629], and the version's version number.

      • Set status to latest_version_at_issuance if the trust store version is currently the latest version. Otherwise, set it to previous_version.

      • Set labels to the trust store entry's labels.

If the above procedure outputs a TrustStoreInclusion with status of latest_version_at_issuance, the certification path's lifetime MUST NOT exceed the corresponding trust store entry's max_lifetime field. This ensures the procedures in Section 6.3 and Section 6.5 interoperate correctly.

The CA combines the outputs of this procedure for each known manifest to form the final TrustStoreInclusionList. To aid the CA in collecting manifests, repositories such as [CCADB] can publish a directory of trust store manifests from participating root programs.

5.3. Media Type

A certification path with its associated CertificationPropertyList may be represented in a PEM [RFC7468] structure in a file of type "application/pem-certificate-chain-with-properties". Files of this type MUST use the strict encoding and MUST NOT include explanatory text. The ABNF [RFC5234] for this format is as follows, where "stricttextualmsg" and "eol" are as defined in Section 3 of [RFC7468]:

certchainwithproperties = stricttextualmsg eol stricttextualmsg
                          *(eol stricttextualmsg)

The first element MUST be the encoded CertificatePropertyList. The second element MUST be an end-entity certificate. Each following certificate MUST directly certify the one preceding it. The certificate that represents the trust anchor MUST be omitted from the path.

CertificatePropertyLists are encoded using the "CERTIFICATE PROPERTIES" label. The encoded data is a serialized CertificatePropertyList, defined in Section 5.

Certificates are encoded as in Section 5.1 of [RFC7468], except DER [X690] MUST be used.

The following is an example file with a certification path containing an end-entity certificate and an intermediate certificate.

-----BEGIN CERTIFICATE PROPERTIES-----
TODO fill in an example
-----END CERTIFICATE PROPERTIES-----
-----BEGIN CERTIFICATE-----
TODO fill in an example
-----END CERTIFICATE-----
-----BEGIN CERTIFICATE-----
TODO fill in an example
-----END CERTIFICATE-----

The IANA registration for this media type is described in Section 12.2.

6. TLS Certificate Negotiation

This section defines the trust_expressions extension, which is sent in the ClientHello, CertificateRequest, and Certificate messages.

6.1. Trust Expressions

When the trust_expressions extension is sent in ClientHello and CertificateRequest, the extension_data field is a TrustExpressionList, defined below:

enum { trust_expressions(TBD), (2^16-1) } ExtensionType;

struct {
    TrustStore trust_store;
    uint24 excluded_labels<0..2^16-1>;
} TrustExpression;

TrustExpression TrustExpressionList<1..2^16-1>;

A TrustExpressionList is an unordered set of TrustExpression objects. When a relying party sends a TrustExpressionList, it indicates support for all trust anchors specified by any TrustExpression contained in the list. A TrustExpression specifies a list of trust anchors in two parts:

First, trust_store specifies a trust store by name and version (see Section 4). Any trust anchors contained in the specified trust store are included.

Second, excluded_labels specifies a set of labels, each of which identify one or more trust anchors in a trust store manifest. Any trust anchors identified by any label in excluded_labels are excluded. Section 6.5 discusses this set in more detail.

excluded_labels MUST be sorted in ascending order and contain no duplicate values to be valid. If invalid, receivers MUST terminate the connection with an illegal_parameter alert.

Together, a TrustExpression indicates that the relying party accepts all trust anchors in the referenced trust store version, except for any that were excluded via excluded_labels.

6.2. Subscriber Acknowledgement

When the trust_expressions extension is sent in a Certificate message, the extension MUST be in the first CertificateEntry and the extension_data field MUST be empty. This extension indicates the subscriber selected a certification path using trust expressions.

In this case, the certificate_list flexibility described in Section 4.4.2 of [RFC8446] no longer applies. The certificate_list MUST contain a complete certification path, correctly ordered and with no extraneous certificates. That is, each certificate MUST certify the one immediately preceding it, and the trust anchor MUST certify the final certificate.

If a relying party receives this extension in the Certificate message, it SHOULD NOT use path building [RFC4158] to validate the result.

6.3. Evaluating Trust Expressions

Given a certification path with a trust_store certificate property (Section 5.1), a subscriber can evaluate a TrustExpressionList to determine whether the certification path matches.

A TrustExpression is said to match a TrustStoreInclusionList if there is at least one TrustStoreInclusion in the TrustStoreInclusionList such that all of the following are true:

  • The name fields of two structures' trust_store fields are equal.

  • Either the version fields of the two structures' trust_store fields are equal, or the TrustStoreInclusion's status is latest_version_at_issuance and the version field of TrustExpression's trust_store is greater than that of the TrustStoreEntry's trust_store.

  • There is no value in common between the TrustStoreInclusion's labels field and the TrustExpression's excluded_labels field.

The invariants in Section 5.1 imply that at most one TrustStoreInclusion will satisfy the first two properties. Implementations may evaluate this efficiently by performing a binary search over the TrustStoreInclusionList, and then checking the third property.

A TrustExpressionList is said to match a TrustStoreInclusionList if any TrustExpression in the TrustExpressionList matches the TrustStoreInclusionList.

When a TrustStoreInclusion's status is latest_version_at_issuance, the above criteria predicts that trust anchors in the latest version will continue to be present in later versions. This allows relying parties using later trust store versions to continue to interoperate with certification paths that predate the later version. If this prediction is incorrect and the trust anchor has been removed from the later version, Section 6.5 requires that the TrustExpression include appropriate excluded_labels values to mitigate this.

6.4. Subscriber Behavior

Subscribers using this negotiation scheme are configured with a list of certification paths, with corresponding CertificatePropertyList (Section 5) structures, in some preference order. When responding to a ClientHello (as a server) or CertificateRequest (as a client) containing the trust_expressions extension, the subscriber collects all candidate certification paths such that all of the following are true:

  • The certification path has not expired.

  • The CertificatePropertyList has a trust_store property with a TrustStoreInclusionList.

  • The matching algorithm described in Section 6.3 returns true.

  • The certification path is suitable for use based on other TLS criteria. For example, the TLS signature_algorithms (Section 4.2.3 of [RFC8446]) extension constrains the types of keys which may be used.

  • The certification path satisfies any other application-specific constraints which may apply. For example, TLS servers often select certificates based on the server_name (Section 3 of [RFC6066]) extension.

Once all candidate paths are determined, the subscriber picks one to present to the relying party. The subscriber MUST include an empty trust_expressions extension in the first CertificateEntry. If multiple candidates match, the subscriber picks its most preferred option. For example, it may try to minimize message size, or prefer options with more performant private keys.

If no candidates match, or if the peer did not send a trust_expressions extension, the subscriber falls back to preexisting behavior outside the scope of this document, without including a trust_expressions extension. For example, the subscriber may have a default certificate configured, or select a certificate using the certificate_authorities extension.

6.5. Constructing Trust Expressions

Relying parties send the trust_expressions extension to advertise a set of accepted trust anchors. Each trust expression to be sent advertises a subset of some trust store version. Relying parties MAY send multiple trust expressions, each referencing a different trust store, if it wishes to communicate a set of trust anchors that span multiple trust stores.

For each referenced trust store version, the following procedure constructs a trust expression. This procedure is expected to be run by the root program, as part of defining new trust store versions and provisioning supported trust anchors in relying parties.

[[TODO: This is written as a procedure, but we expect root programs to fill in the expected exclusions when they define a new trust store version, and then trim the compatibility exclusions as they expire. Also the root programs know their label allocation scheme and are the ones deciding on removals, so they're best situated to pick a set of excluded_labels. Perhaps this should get moved to the manifest?]]

  1. Let include_entries and exclude_entries be two empty sets of trust anchor entries.

  2. For each trust store entry in the trust store version:

    1. If the trust store entry's id references a trust anchor that is in the desired subset, add it to include_entries.

    2. Otherwise, add it to exclude_entries.

  3. For all trust store entries in trust store versions before the specified version:

    1. If the current time is before the entry's expiration time (Section 4.1) and if the entry's id references a trust anchor that is not in the desired subset, add the entry to exclude_entries.

  4. Compute a set of labels, excluded_labels such that:

    1. No label appears in any entry in include_entries.

    2. For each entry in exclude_entries, there is some label in common between excluded_labels and the labels in the entry.

    How to compute this set depends on the root program's label allocation scheme (Section 4.2). If the root program allocates a label for each trust anchor, this set is always computable, though more efficient sets may be possible depending on the allocation scheme.

  5. Output a TrustExpression such that trust_store is the referenced trust store version, and excluded_labels is the computed excluded_labels value.

This procedure uses excluded_labels for two kinds of exclusions:

First, if the trust store version includes undesired trust anchors, the trust expression should exclude them. This may occur if, for example, the trust store is used by all versions of the relying party's software, but some trust anchors are gated by software version.

Second, trust expressions exclude unexpired entries from previous versions. This is because the matching criteria described in Section 6.3 predictively applies TrustStoreEntry values with status of latest_version_at_issuance to all future versions of a trust store. This allows relying parties to interoperate with subscribers with stale information. Unexpired entries are those for which such an unexpired certification path may still exist. Where this prediction is incorrect, trust expressions MUST mitigate this by excluding the past entries.

7. Issuing Certificates

Subscribers SHOULD use an automated issuance process where the CA transparently provisions multiple certification paths, without changes to subscriber configuration. As relying party requirements evolve, the CA adjusts its output to ensure its subscribers continue to interoperate with all supported relying parties. This results in a more flexible and agile PKI that can better respond to security incidents and changes in requirements. (See Section 9 for details.)

Subscribers MAY additionally combine the outputs from multiple CAs. This may be used, for example, to maintain backup certificates as described in Section 9.6.

7.1. ACME Extension

This section extends ACME [RFC8555] to issue multiple certification paths, each with a CertificatePropertyList, within a single ACME order. It extends the ACME order object with a configurable boolean field, "trustExpressions".

First, ACME clients indicate they support certification paths selected by trust expressions by setting "trustExpressions" to "true" in the ACME newOrder request (Section 7.4 of [RFC8555]). If the server accepts the request, it MUST reflect the field in the resulting order object.

When an ACME server processes an order object with the "trustExpressions" field set to "true", it MAY issue multiple certification paths, each with an associated TrustStoreInclusionList, computed as in Section 5.2. These paths MAY share an end-entity certificate but end at different trust anchors, or they MAY be completely distinct. If the latter, the "trustExpressions" order field allows the ACME server to skip issuing unnecessary end-entity certificates for ACME clients that would be unable to use them. The ACME server encodes each certification path using the application/pem-certificate-chain-with-properties format, defined in Section 5.3). Note this format is required to form a complete certification path. The CA MUST return a result that may be verified by relying parties without path building [RFC4158].

The ACME server provides additional results to the client with the link relation header fields described in Section 7.4.2 of [RFC8555]. When fetching certificates, the ACME client includes application/pem-certificate-chain-with-properties in its Accept header to indicate it supports the new format. Unlike the original mechanism described in [RFC8555], these certification paths do not require heuristics to be used. Instead, the ACME client uses the associated TrustStoreInclusionLists to select a path as described in Section 6.4.

When the ACME client wishes to renew any of the certification paths issued in this order, it repeats this process to renew each path concurrently. Thus this extension is suitable when the CA is willing to issue and renew all paths together. It may not be appropriate if the paths have significantly different processing times or lifetimes. Future enhancements to ACME may be defined to address these cases, e.g. by allowing the ACME client to make independent orders.

8. Example

Suppose A1, A2, B1, B2, C1, and C2 are X.509 root certificates, such that A1 and A2 share an operator, B1 and B2 share an operator, and C1 and C2 share an operator.

On January 1st, 2023, the root program includes A1, A2, B1, and B2. It allocates labels 0 through 3 for each individual CA and then 100 and 101 for the two CA operators. The CAs issue 90 day certificates and are permitted to use manifests stale by 10 days. The root store manifest may then be:

{
  "name": "example",
  "max_age": 864000,
  "trust_anchors": {
    "A1": {"type": "x509", "data": "..."},
    "A2": {"type": "x509", "data": "..."},
    "B1": {"type": "x509", "data": "..."},
    "B2": {"type": "x509", "data": "..."}
  },
  "versions": [
    {
      "timestamp": 1672531200,
      "entries": [
        {"id": "A1", "labels": [0, 100],
         "max_lifetime": 7776000},
        {"id": "A2", "labels": [1, 100],
         "max_lifetime": 7776000},
        {"id": "B1", "labels": [2, 101],
         "max_lifetime": 7776000},
        {"id": "B2", "labels": [3, 101],
         "max_lifetime": 7776000}
      ]
    }
  ]
}

A certification path, A1_old, issued by A1 would then contain the TrustStoreEntry:

A certification path, B1_old, issued by B1 would then contain the TrustStoreEntry:

A certification path, C1_old, issued by C1 would contain no TrustStoreEntry values that reference "example".

On February 1st, 2023, the root program added CAs C1 and C2 but removed CAs B1 and B2. It continues the previous label allocation scheme, but now wishes to allocate label 200 for CAs A1 and C1. The manifest may then be:

{
  "name": "example",
  "max_age": 864000,
  "trust_anchors": {
    "A1": {"type": "x509", "data": "..."},
    "A2": {"type": "x509", "data": "..."},
    "B1": {"type": "x509", "data": "..."},
    "B2": {"type": "x509", "data": "..."},
    "C1": {"type": "x509", "data": "..."},
    "C2": {"type": "x509", "data": "..."}
  },
  "versions": [
    {
      "timestamp": 1672531200,
      "entries": [
        {"id": "A1", "labels": [0, 100],
         "max_lifetime": 7776000},
        {"id": "A2", "labels": [1, 100],
         "max_lifetime": 7776000},
        {"id": "B1", "labels": [2, 101],
         "max_lifetime": 7776000},
        {"id": "B2", "labels": [3, 101],
         "max_lifetime": 7776000}
      ]
    },
    {
      "timestamp": 1675209600,
      "entries": [
        {"id": "A1", "labels": [0, 100, 200],
         "max_lifetime": 7776000},
        {"id": "A2", "labels": [1, 100],
         "max_lifetime": 7776000},
        {"id": "C1", "labels": [4, 102, 200],
         "max_lifetime": 7776000},
        {"id": "C2", "labels": [5, 102],
         "max_lifetime": 7776000}
      ]
    }
  ]
}

A certification path, A1_new, now issued by A1 would contain two TrustStoreEntry values. The first:

And the second:

A certification path, B1_new, now issued by B1 would contain a TrustStoreEntry:

A certification path, C1_new, now issued by C1 would contain a TrustStoreEntry:

A relying party which trusts trust anchors A1, A2, B1, and B2 might send a TrustExpression referencing trust store "example", version 0, with empty excluded_labels. This would match A1_old, A1_new, B1_old, and B1_new by the corresponding TrustStoreEntry. It would not match C1_old or C1_new.

A relying party which trusts trust anchors A2, B1, and B2, but not A1, might send a TrustExpression referencing trust store "example", version 0, with excluded_labels of 0. This would match B1_old and B1_new, but not A1_old or A1_new because the TrustExpression excludes A1.

A relying party which trusts trust anchors A1, A2, C1, and C2 might send a TrustExpression referencing trust store "example", version 1, with excluded_labels of 101. Although B1 and B2 are not contained in version 1 of the trust store, the relying party must exclude them, per Section 6.5. This TrustExpression matches the above certification paths as follows:

The relying party could equivalently have sent a TrustExpression referencing trust store "example", version 1, with excluded_labels of 2 and 3. This pair of labels excludes the same CAs as 101.

Of the above examples, B1_old depends on the exclusion to avoid a false positive match. 100 days after February 1st, B1 and B2's trust store entries from version 0 will expire (see Section 4.1), and the relying party no longer needs to exclude them. By this point, B1_old will have expired. B1_new may still be valid, but it does not require the exclusion.

9. Use Cases

9.1. Key Rotation

In most X.509 deployments a compromise of any root CA's private key would compromise the entire PKI. Despite this, X.509 deployments often do not rotate root keys. The oldest certificate in [CHROME-ROOTS] and [MOZILLA-ROOTS] was issued in 1998, 25 years old as of writing in 2023. A single-certificate deployment model makes rotating root keys challenging due to the need for that certificate to be trusted by a wide range of relying parties. Not all relying parties may be quickly updated with the new root CA, either due to being offline or simply lacking an update mechanism. This means subscribers cannot switch to serving the new root, which in turn means relying parties cannot distrust the old root to complete the transition.

A multi-certificate deployment model avoids these transition problems. Key rotation may proceed as follows:

  1. The CA operator generates a new root CA with a separate key, but continues operating the old root CA.

  2. Root programs begin trusting the new root CA alongside the old one. They update their root store manifests.

  3. When subscribers request certificates, the CA issues certificates from both roots and provisions the subscriber with both certificates.

  4. Relying parties send trust expressions that evaluate to either the old or new root, and are served the appropriate option.

  5. Once subscribers have been provisioned with new certificates, root programs can distrust the old root without causing relying party validation failures. The CA operator continues to operate the old root CA for as long as it wishes to serve subscribers that, in turn, wish to serve older relying parties.

Moreover, this process can complete with no configuration changes to the subscriber, given an automated, multi-certificate-aware certificate issuance process. The subscriber does not need to know why it received two certificates, only how to select between them for each relying party, as specified in Section 6.4.

9.2. Adding CAs

In the single-certificate X.509 model widely deployed today, subscribers cannot use TLS certificates issued from a new root CA until all supported relying parties have been updated to trust the new root CA. In practice, it can take several years for a new root CA to become trusted in browsers. Some relying parties, such as IoT devices, may never receive trust store updates at all.

As a result, it is very difficult for subscribers that serve a wide variety of relying parties to use a newly-trusted root CA. When trust stores diverge too far, subscribers often must partition their services into multiple TLS endpoints (i.e. different DNS names) and direct different relying parties to different endpoints. Subscribers sometimes resort to TLS fingerprinting, to detect particular relying parties. But, as this repurposes other TLS fields for unintended purposes, this is unreliable and usually requires writing custom service-specific logic.

In a multi-certificate deployment model, subscribers can begin serving certificates from new root CAs without interrupting relying parties that depend on existing ones.

In some contexts, it may be possible to use other fields to select the new CA. For example, post-quantum-capable clients may be detected with the signature_algorithms and signature_algorithms_cert extensions. However, this assumes all post-quantum CAs are added at the same time. A multi-certificate model avoids this problem and allows for a more gradual deployment of post-quantum CAs.

9.3. Removing CAs

To serve certificates that will validate in all supported relying parties, subscribers in a single-certificate model only serve certificates from CAs in the intersection of multiple relying party trust stores. As relying parties remove untrusted CAs over time, this intersection may shrink, reducing the set of CAs available to subscribers that can issue widely-trusted certificates. Two relying parties may each support many CAs, but may have no common CA in common.

This is exacerbated by relying parties that either have not or cannot take updates, and so cannot trust newer CAs. Moreover, as relying parties do not currently report their trust stores, the subscriber may not even know which CAs are in the intersection. Often, the only option is to try the new certificate and monitor errors. For subscribers that serve many diverse relying parties, switching CAs is a disruptive and risky process.

In a multi-certificate model, there is no need for subscribers to limit themselves to this intersection. In a multi-certificate model, a subscriber can obtain certificates from multiple CAs, such that each supported relying party is covered by some certificate in the set.

For example, suppose a subscriber uses certificates from some CA, CA1. However, some relying party no longer trusts CA1. The subscriber can obtain certificates from another CA that is trusted by that relying party, CA2, while continuing to also use CA1. The relying party that no longer trusts CA1 will be served the CA2's certificate, while relying parties that still trust CA1 can be served the original one.

9.4. Intermediate Elision

In many PKIs today, root CAs issue certificates to intermediate CAs which, in turn, issue end-entity certificates. This gives some operational and security improvements over issuing end-entity certificates directly from a root CA. The root certificate's private key is less exposed to attack and it allows for different issuance and ownership models. While the intermediate CA often inherits all the authority of the root CA, it can be replaced without changes to relying parties. Thus it can be more easily revoked and/or have a shorter lifetime.

However, this comes at a size cost: the certification path, sent in the TLS handshake, includes an extra certificate. This is an extra public key and signature, as well as extra copies of X.509 metadata. (An average X.509 name in the Chrome Root Store [CHROME-ROOTS] or Mozilla CA Certificate Program [MOZILLA-ROOTS] is around 100 bytes, despite largely being an opaque identifier in modern usage.) Post-quantum signature algorithms shift this tradeoff dramatically. Dilithium3 [Dilithium], for example, has a total public key and signature size of 5,245 bytes.

[I-D.ietf-tls-cert-abridge] proposes to predistribute known intermediate certificates to relying parties, as a compression scheme. A multi-certificate deployment model provides another way to achieve this effect. To relying parties, a predistributed intermediate certificate is functionally equivalent to a root certificate. PKIs use intermediate certificates because changing root certificates requires updating relying parties, but predistributed intermediates already presume updated relying parties.

A CA operator could provide subscribers with two certification paths: a longer path ending at a long-lived trust anchor and shorter path the other ending at a short-lived, revocable root. Relying parties would be configured to trust both the long-lived root and the most recent short-lived root. The negotiation mechanism in Section 6 would then send the shorter path to up-to-date relying parties, and the longer path to older relying parties.

This achieves the same effect with a more general-purpose multi-certificate mechanism. It is also more flexible, as the two paths need not be related. For example, root CA public keys are not distributed in each TLS connection, so a post-quantum signature algorithm that optimizes for signature size may be preferable. In this model, both the long-lived and short-lived root may use such an algorithm. In a compression-based model, the same intermediate must optimize both its compressed and uncompressed size, so such an algorithm may not be suitable.

9.5. Conflicting Relying Party Requirements

A subscriber may need to support relying parties with different requirements. For example, in contexts where online revocation checks are expensive, unreliable, or privacy-sensitive, user security is best served by short-lived certificates. In other contexts, long-lived certificates may be more appropriate for, e.g., systems that are offline for long periods of time or have unreliable clocks.

A single-certificate deployment model forces subscribers to find a single certificate that meets all requirements. User security then suffers in all contexts, as the PKI may not quite meet anyone's needs. In a multi-certificate deployment model, different contexts may use different trust anchors. A subscriber that supports multiple contexts would provision certificates for each, with certificate negotiation logic directing the right one to each relying party.

9.6. Backup Certificates

A subscriber desiring to offer a robust service to relying parties in the face of a potential future CA compromise or distrust may obtain certification paths from multiple CAs. This provides redundancy in the event that some relying party no longer supports one of the two CAs. As long as the subscriber continues to have at least one trusted certification path, it will continue to be trusted by the relying parties.

In order to support this, TLS serving software SHOULD permit users to configure multiple ACME endpoints and select from the the union of the certificate paths returned by each ACME server.

10. Privacy Considerations

The negotiation mechanism described in this document presumes the relying party's trust anchor list is not sensitive. While an attacker may already query this list by probing which certification paths are accepted by the relying party, this extension introduces a faster and more reliable method.

Thus, this mechanism SHOULD NOT be used in contexts where the list reveals information about an individual user. For example, a web browser may support both a common set of trust anchors configured by the browser vendor, and a set of user-specified trust anchors. The common trust anchors would only reveal which browser is used, while the user-specified trust anchors may reveal identifying information about the user. In this case, the trust anchor list SHOULD be limited to the common trust anchors. This limits the benefits of trust anchor agility to subscribers that use one of the advertised trust anchors. However, deployments that rely on an unadvertised trust anchor will fall back to the default behavior, as if this extension were not used.

Likewise, a client that allows user-specified distrusts of common trust anchors will reveal user-specific information if distrusts are reflected in transmitted trust expressions. Trust expressions SHOULD be constructed as if those trust anchors were trusted. This may result in servers selecting a distrusted certificate, despite a trusted one being available.

Additionally, two relying parties may express the same trust anchor list with different trust expressions. Doing so may also reveal information about an individual user. Relying parties within an anonymity set (Section 3.3 of [RFC6973]) SHOULD send the same trust expression. To achieve this, the trust expressions SHOULD be assembled by the root program and configured in relying parties alongside trust store updates.

11. Security Considerations

The certificate negotiation mechanism described in this document facilitates which certification path is served to relying parties, but has no impact on the relying party’s trust preferences themselves.

As a result, this allows for a more flexible and agile PKI, which can better mitigate security risks to users. Section 9 discusses some scenarios which benefit from the multi-certificate deployment this document enables. In general, robust certificate negotiation helps subscribers navigate differences in relying party requirements. This means security improvements for one set of relying parties can be deployed without needing to risk incompatibility or breakage for others.

If either the subscriber's TrustStoreInclusionList or the relying party's TrustExpressionList are incorrect, the matching algorithm described in Section 6.4 may incorrectly identify an untrusted certification path. This mechanism will not result in that path being trusted, but does present the possibility of a denial of service. These structures are provisioned by the CA and root program, respectively, who are already expected to provide accurate information.

12. IANA Considerations

12.1. TLS ExtensionType Updates

IANA is requested to create the following entry in the TLS ExtensionType Values registry, defined by [RFC8446]:

Table 1
Value Extension Name TLS 1.3 DTLS-Only Recommended Reference
TBD trust_expressions CH, CR, CT N Y [this-RFC]

12.2. Media Type Updates

IANA is requested to create the following entry in the "Media Types" registry, defined in [RFC6838]:

Type name:

application

Subtype name:

pem-certificate-chain-with-properties

Required parameters:

None

Optional parameters:

None

Encoding considerations:

7bit

Security considerations:

Carries a cryptographic certificate and its associated certificate chain and additional properties. This media type carries no active content.

Interoperability considerations:

None

Published specification:

[this-RFC, Section 5.3]

Applications that use this media type:

ACME clients and servers, HTTP servers, other applications that need to be configured with a certificate chain

Additional information:
Deprecated alias names for this type:
n/a
Magic number(s):
n/a
File extension(s):
.pem
Macintosh file type code(s):
n/a
Person & email address to contact for further information:

See Authors' Addresses section.

Intended usage:

COMMON

Restrictions on usage:

n/a

Author:

See Authors' Addresses section.

Change controller:

IESG

12.3. ACME Order Object Fields Updates

IANA is requested to create the following entry in the ACME Order Object Fields registry, defined by [RFC8555]:

Table 2
Field Name Field Type Configurable Reference
trustExpressions boolean true [this-RFC]

12.4. CertificatePropertyType Registry

[[TODO: Establish a CertificatePropertyType registry.]]

13. CDDL Schema

The following is a non-normative CDDL [RFC8610] schema which describes a trust store manifest structure (Section 4):

trust-anchor = {
    type: text,
    * text => any,
}

trust-store-entry = {
    id: text,
    labels: [+ uint],
    max_lifetime: uint,
    * text => any,
}

trust-store-version = {
    timestamp: int,
    entries: [+ trust-store-entry],
    * text => any,
}

trust-store-manifest = {
    name: text,
    max_age: uint,
    trust_anchors: {+ text => trust-anchor},
    versions: [+ trust-store-version],
    * text => any,
}

Acknowledgements

The authors thank Nick Harper, Sophie Schmieg, and Emily Stark for many valuable discussions and insights which led to this document, as well as review of early iterations.

References

Normative References

[POSIX]
"IEEE Standard for Information Technology--Portable Operating System Interface (POSIX(TM)) Base Specifications, Issue 7", IEEE, DOI 10.1109/ieeestd.2018.8277153, ISBN ["9781504445429"], , <https://doi.org/10.1109/ieeestd.2018.8277153>.
[RFC2119]
Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, , <https://www.rfc-editor.org/rfc/rfc2119>.
[RFC3629]
Yergeau, F., "UTF-8, a transformation format of ISO 10646", STD 63, RFC 3629, DOI 10.17487/RFC3629, , <https://www.rfc-editor.org/rfc/rfc3629>.
[RFC4158]
Cooper, M., Dzambasow, Y., Hesse, P., Joseph, S., and R. Nicholas, "Internet X.509 Public Key Infrastructure: Certification Path Building", RFC 4158, DOI 10.17487/RFC4158, , <https://www.rfc-editor.org/rfc/rfc4158>.
[RFC4648]
Josefsson, S., "The Base16, Base32, and Base64 Data Encodings", RFC 4648, DOI 10.17487/RFC4648, , <https://www.rfc-editor.org/rfc/rfc4648>.
[RFC5234]
Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax Specifications: ABNF", STD 68, RFC 5234, DOI 10.17487/RFC5234, , <https://www.rfc-editor.org/rfc/rfc5234>.
[RFC5280]
Cooper, D., Santesson, S., Farrell, S., Boeyen, S., Housley, R., and W. Polk, "Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 5280, DOI 10.17487/RFC5280, , <https://www.rfc-editor.org/rfc/rfc5280>.
[RFC6838]
Freed, N., Klensin, J., and T. Hansen, "Media Type Specifications and Registration Procedures", BCP 13, RFC 6838, DOI 10.17487/RFC6838, , <https://www.rfc-editor.org/rfc/rfc6838>.
[RFC6973]
Cooper, A., Tschofenig, H., Aboba, B., Peterson, J., Morris, J., Hansen, M., and R. Smith, "Privacy Considerations for Internet Protocols", RFC 6973, DOI 10.17487/RFC6973, , <https://www.rfc-editor.org/rfc/rfc6973>.
[RFC7468]
Josefsson, S. and S. Leonard, "Textual Encodings of PKIX, PKCS, and CMS Structures", RFC 7468, DOI 10.17487/RFC7468, , <https://www.rfc-editor.org/rfc/rfc7468>.
[RFC8174]
Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, , <https://www.rfc-editor.org/rfc/rfc8174>.
[RFC8259]
Bray, T., Ed., "The JavaScript Object Notation (JSON) Data Interchange Format", STD 90, RFC 8259, DOI 10.17487/RFC8259, , <https://www.rfc-editor.org/rfc/rfc8259>.
[RFC8446]
Rescorla, E., "The Transport Layer Security (TLS) Protocol Version 1.3", RFC 8446, DOI 10.17487/RFC8446, , <https://www.rfc-editor.org/rfc/rfc8446>.
[RFC8555]
Barnes, R., Hoffman-Andrews, J., McCarney, D., and J. Kasten, "Automatic Certificate Management Environment (ACME)", RFC 8555, DOI 10.17487/RFC8555, , <https://www.rfc-editor.org/rfc/rfc8555>.
[X690]
ITU-T, "Information technology - ASN.1 encoding Rules: Specification of Basic Encoding Rules (BER), Canonical Encoding Rules (CER) and Distinguished Encoding Rules (DER)", ISO/IEC 8825-1:2002, .

Informative References

[CCADB]
Mozilla, Microsoft, Google, Apple, and Cisco, "Common CA Database", , <https://www.ccadb.org/>.
[CHROME-ROOTS]
Chromium, "Chrome Root Store", , <https://chromium.googlesource.com/chromium/src/+/main/net/data/ssl/chrome_root_store>.
[Dilithium]
Bai, S., Ducas, L., Kiltz, E., Lepoint, T., Lyubashevsky, V., Schwabe, P., Seiler, G., and D. Stehlé, "CRYSTALS-Dilithium Algorithm Specifications and Supporting Documentation", , <https://pq-crystals.org/dilithium/data/dilithium-specification-round3-20210208.pdf>.
[I-D.ietf-tls-cert-abridge]
Jackson, D., "Abridged Compression for WebPKI Certificates", Work in Progress, Internet-Draft, draft-ietf-tls-cert-abridge-00, , <https://datatracker.ietf.org/doc/html/draft-ietf-tls-cert-abridge-00>.
[MOZILLA-ROOTS]
Mozilla, "Mozilla Included CA Certificate List", , <https://wiki.mozilla.org/CA/Included_Certificates>.
[RFC6066]
Eastlake 3rd, D., "Transport Layer Security (TLS) Extensions: Extension Definitions", RFC 6066, DOI 10.17487/RFC6066, , <https://www.rfc-editor.org/rfc/rfc6066>.
[RFC8610]
Birkholz, H., Vigano, C., and C. Bormann, "Concise Data Definition Language (CDDL): A Notational Convention to Express Concise Binary Object Representation (CBOR) and JSON Data Structures", RFC 8610, DOI 10.17487/RFC8610, , <https://www.rfc-editor.org/rfc/rfc8610>.

Authors' Addresses

David Benjamin
Google LLC
Devon O'Brien
Google LLC
Bob Beck
Google LLC