RFC 5697

Other Certificates Extension 🎀

Okay bestie, RFC 5697 is the Other Certificates Extension and it's giving experimental research energy 🔬✨ It's by S. Farrell at Trinity College Dublin, published under the Network Working Group in November 2009, and its category is Experimental, so it's a "try it and tell us how it goes" kind of spec 💅 The header has no Obsoletes or Updates line, so it doesn't replace or change any other RFC.

Abstract ✨

Okay so some apps attach state info to public key certificates and they'd be SO much happier if they could link together a bunch of certs that belong to the same end entity and can safely be treated as equivalent for looking up that state 🫶 This memo defines a certificate extension that lets apps set up that linkage without needing a brand-new application protocol data unit 🦋

Status of This Memo 💜

This one is an Experimental Protocol for the Internet community, so it's not an Internet standard of any kind 🙈 They're asking for discussion and suggestions for improvement, and distribution of this memo is unlimited.

Copyright Notice 🌈

Copyright (c) 2009 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 (http://trustee.ietf.org/license-info) as they stand on the date of publication. Please read these carefully, because they describe your rights and restrictions for this document. Code components pulled out of this document must include Simplified BSD License text, as described in Section 4.e of the Trust Legal Provisions, and they come without warranty as described in the BSD License 📜

Table of Contents 📚

1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.  A Use Case  . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.  Other Certificates Extension  . . . . . . . . . . . . . . . . . 3
4.  Another Approach Using Permanent Identifiers  . . . . . . . . . 5
5.  A Possible Optimisation . . . . . . . . . . . . . . . . . . . . 5
6.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . 6
7.  Security Considerations . . . . . . . . . . . . . . . . . . . . 6
8.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . 7
  8.1.  Normative References  . . . . . . . . . . . . . . . . . . . 7
  8.2.  Informative References  . . . . . . . . . . . . . . . . . . 7
Appendix A.  ASN.1 Module . . . . . . . . . . . . . . . . . . . . . 8

1. Introduction ✨

RFC 5280 [RFC5280] is the profile for using public key certificates on Internet applications. If an app associates application-state information with a certificate, that association can get disrupted when the end entity changes its certificate, whether that's because of a renewal or because it switched certificate issuer 😭 Also, if the end entity is actually a distributed system where each instance has its own private key, the relying party (RP) has no way to connect the different certificates to the right application-state information 🧚

Here's an example: say a web browser keeps state info (maybe passwords) about a website, indexed (possibly indirectly) through values in the web server's public key certificate (maybe a DNS name). When that web server certificate expires and a new one is acquired (maybe with a different DNS name), the browser can't safely map the new certificate to the state it has saved 💀

This memo defines a new public key certificate extension that supports this kind of linkage. It lets the certificate issuer attest that the end entity holding the private key for the certificate in question also holds other private keys corresponding to other identified certificates 🔥

Other than the issuer asserting that a set of certificates belongs to the same end entity for use with the same application, the fine detail of how certificate linkage works isn't defined here. That's up to application developers and the operators of certification authorities (CAs) 🌸 In particular, we don't define how a CA can validate that the same end entity holds the various private keys, or how an application should use this info. We also don't define what kinds of state information may be shared.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119] 🫶

2. A Use Case 🦋

Public key certificates expire, typically about a year after they're created. Some applications might need to know that the same entity is the subject of the current certificate and a certificate used previously 🌈

For example, if a web server certificate expires, it could be useful for a web browser to know that the server presenting a certificate during a Transport Layer Security (TLS) [RFC5246] handshake is the same web server that presented a certificate before. That could let the browser automatically fill in form fields for that server, even if the server certificate has been replaced 💖 You can get the same effect by using the same issuer and subject fields in a certificate, but there could be security issues with those comparisons, e.g., if the subject name includes a DNS name and the ownership of that DNS domain has changed 🙈

The new extension gives the CA a way to signal to the application that the same end entity is involved, no matter what name changes happened. It could also make it easier for a web site operator to change CA when replacing its certificate 👑

3. Other Certificates Extension 💅

This section defines the syntax for the other certificates extension.

The new extension is basically a list of references to the linked certificates. The references use the SCVPCertID structure from the Server-Based Certificate Validation Protocol (SCVP) [RFC5055], which contains a hash over the relevant certificate and the certificate's issuer and serial number 🧠

When this extension is present, the CA is asserting that the same end entity is the subject of the relevant certificates ✨

This extension MUST NOT be marked critical.

id-pe-otherCerts OBJECT IDENTIFIER ::= { id-pe 19 }
OtherCertificates ::= SEQUENCE OF SCVPCertID

CAs MUST only issue certificates containing this extension where the links they create are such that the relevant consumers of the certificates can safely use those links. That's typically the case where the certificates are only used by a single application 🎀 CAs MUST NOT issue certificates that link to certificates issued for a different purpose. For example, a CA SHOULD NOT link a web server certificate to a VPN gateway certificate (unless those can be the same, which might happen for some embedded devices). The purpose a certificate is intended for may be determined by certificate policy or other means (e.g., extended key usage object identifiers), and those are out of scope for this specification 🌟

CAs MUST NOT issue certificates containing this extension unless they have validated that the end entity holds all of the relevant private keys 🔐

Applications MUST validate certificates according to the rules in RFC 5280 [RFC5280], and they MUST NOT assume that just because certificates are linked, they're therefore valid. This means, obviously, that both certificates must chain up to some local trust point(s) 🚨

If an application adds further checks on certificate validity (for example, as RFC 2818 [RFC2818] does for web server certificates), then both certificates MUST be valid according to those application-specific rules 💜

It is not required that two linked certificates be valid at the same time. For example, an application can validate certificate1 and cache that information. When the application is later presented with certificate2 (linked back to certificate1), if it considers the cached information about certificate1 trustworthy, then it can validate certificate2 and use the linkage to associate certificate2 with the relevant application-state information. That's exactly what it would have done if certificate1 had been presented again 🦋 As a second example, if certificate1 has expired but would otherwise be valid, the linkage from certificate2 can still be used once certificate2 has been validated 🌸

If the application checks certificate status for the certificates in question and any of them has been revoked, then the linkage MUST NOT be used 🙅‍♀️

Note that there are no constraints on the contents of the certificate the link points to. So a CA issuing a new certificate can link back to legacy certificates of all kinds, once the relevant RP supports this extension 🧚

This extension MUST only be used in end-entity certificates. In other words, it MUST NOT be used in CA certificates or other similar certificates. CA certificates are only used for certificate validation, and this extension has no effect on the validation procedure, so the extension would generally be meaningless in a CA certificate 🤷‍♀️ It may also be wise to gain some deployment experience with this extension before using it for more security-sensitive certificates, like CA certificates 🫶

4. Another Approach Using Permanent Identifiers 💖

RFC 4043 [RFC4043] defines a new name form (a "Permanent Identifier" or PI) for public key certificates that supports functionality similar to the new extension defined here. If two certificates have the same PI and that PI form is globally unique, then the end entities involved can be considered to be the same 🌈

The main difference between the PI and the other certificates extension is that PI, when more than one CA is involved, requires a globally unique identifier. The other certificates extension only requires that the issuer of the new certificate be able to link back to the old certificate(s) 👑

As a result, the other certificates extension can be deployed "reactively" to link certificates that may not match "ideal" application-naming requirements. If the old certificate did use a PI, then application-naming issues have presumably already been handled, and the new certificate can contain the same PI. In that latter case, there would be no need for the other certificates extension 🎀

5. A Possible Optimisation ✨

The SCVPCertID structure used here contains the issuer name for the CA of the linked certificate. That issuer can also be the issuer of the certificate containing the other certificates extension. If a new certificate were linked back to a number of old certificates from that same CA, there would be a lot of redundancy, since there would be many copies of the same issuer name 🧠

One suggestion raised was a convention where, if the X.500 Name in the SCVPCertID is an "empty" DN (see RFC5280), that would indicate the same CA issued both the current and the linked certificates. However, that scheme isn't adopted in this version. A future, Standards Track version of this specification might adopt that optimisation 🔮

6. Acknowledgements ✨

okay bestie, big thank yous 🫶 The use case behind all this was brought to the W3C web security context (WSC) working group by Tyler Close, and if you want the full story, the details live at http://www.w3.org/2006/WSC/wiki/SafeWebFormEditor 🌸

Denis Pinkas pointed out that the PI extension is an alternative to this one 💜

James Manger suggested the optimisation that cuts down on the number of copies of the issuer name 👑

---

7. Security Considerations 💀🔐

okay this section is SERIOUS, so read it carefully bestie 😭 As already stated above, relying parties MUST validate any certificates per the algorithm given in RFC 5280 [RFC5280] before making any use of those certificates. No skipping that step, periodt 💅

Relying parties similarly MUST NOT assume that any other fields in the relevant certificates have common values. For example, linked certificates might have non-overlapping key usage extensions. Not every field is gonna match, so don't assume it does 🙈

Here's the vibe. The issuer of the new certificate (or some superior CA) is trusted by the RP, and the RP has validated the new certificate. So the RP is basically as reliant on that CA behaving properly as always. If the CA wanted to "cheat" on the RP, the other certificates extension just gives it a new way to do that, but that way is equivalent to existing vulnerabilities, so nothing brand new there 🥀 In a lot of cases, a sketchy CA could just issue a brand new certificate that's identical in every respect (other than the key pair), and the RP would accept the identity in that new certificate anyway 😬

However, if the issuer of the new certificate is limited in some way (for example, via a name constraint in a superior CA certificate), and the old certificate doesn't match those limits (for example, the subject of the old certificate doesn't fit under the name constraints of the issuer of the new certificate), then the new certificate could be linked back to an identity that doesn't meet the constraints that were meant to apply to the issuer of the new certificate. Yikes 😱 Applications where this is an unacceptable risk SHOULD NOT make use of the other certificates extension 🚫

Since the SCVPCertID structure includes a hash of the other certificate, and hash algorithm weaknesses that produce collisions are becoming more of an issue, CAs and relying parties MUST ensure that currently acceptable hash functions are used. Especially, the default use of SHA-1 for SCVPCertID may or may not currently be considered acceptable. CAs might be wise to use SHA-256 instead, but they'll typically just use whatever hash function they use as part of certificate signing 🔍

In some application contexts, if the old certificate has expired (and maybe any associated certificate revocation list (CRL) entries are no longer on the latest CRL), it may be unsafe to link the new and old certificates. Application developers SHOULD carefully consider whether to make use of the other certificates extension in those contexts 🧐✨

---

8. References 📚

8.1. Normative References

• [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.
• [RFC5055] Freeman, T., Housley, R., Malpani, A., Cooper, D., and W. Polk, "Server-Based Certificate Validation Protocol (SCVP)", RFC 5055, December 2007.
• [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, May 2008.

8.2. Informative References

• [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000.
• [RFC4043] Pinkas, D. and T. Gindin, "Internet X.509 Public Key Infrastructure Permanent Identifier", RFC 4043, May 2005.
• [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security (TLS) Protocol Version 1.2", RFC 5246, August 2008.

---

Appendix A. ASN.1 Module 🧚

PKIX OID registrations may be viewed at: http://www.imc.org/ietf-pkix/pkix-oid.asn

PKIXOtherCertsModule
     { iso(1) identified-organization(3) dod(6) internet(1)
       security(5) mechanisms(5) pkix(7) id-mod(0) 44 }

DEFINITIONS EXPLICIT TAGS ::=

BEGIN

-- EXPORTS ALL

IMPORTS

-- From [RFC5055]
SCVPCertID
FROM SCVP { iso(1) identified-organization(3) dod(6) internet(1)
               security(5) mechanisms(5) pkix(7) id-mod(0) 21 } ;

-- The one and only new thing, a new certificate extension

id-pe-otherCerts OBJECT IDENTIFIER ::=
          { iso(1) identified-organization(3) dod(6) internet(1)
               security(5) mechanisms(5) pkix(7) id-pe(1) 19 }

-- The value is a sequence of cert ids.
OtherCertificates ::= SEQUENCE OF SCVPCertID

END

---

Author's Address 🎀

Stephen Farrell Trinity College Dublin Department of Computer Science Trinity College Dublin, 2 Ireland

Phone: +353-1-896-2354 EMail: stephen.farrell@cs.tcd.ie