tech-invite   World Map     

IETF     RFCs     Groups     SIP     ABNFs    |    3GPP     Specs     Gloss.     Arch.     IMS     UICC    |    Misc.    |    search     info

RFC 2510

Pages: 72
Top     in Index     Prev     Next

Internet X.509 Public Key Infrastructure Certificate Management Protocols

Part 1 of 3, p. 1 to 19
None       Next RFC Part

Obsoleted by:    4210

Top       Page 1 
Network Working Group                                            C. Adams
Request for Comments: 2510                           Entrust Technologies
Category: Standards Track                                      S. Farrell
                                                               March 1999

                Internet X.509 Public Key Infrastructure
                    Certificate Management Protocols

Status of this Memo

   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice

   Copyright (C) The Internet Society (1999).  All Rights Reserved.


   This document describes the Internet X.509 Public Key Infrastructure
   (PKI) Certificate Management Protocols. Protocol messages are defined
   for all relevant aspects of certificate creation and management.
   Note that "certificate" in this document refers to an X.509v3
   Certificate as defined in [COR95, X509-AM].

   The key words "MUST", "MUST NOT", "REQUIRED", "SHOULD", "SHOULD NOT",
   "RECOMMENDED", "MAY", and "OPTIONAL" in this document (in uppercase,
   as shown) are to be interpreted as described in [RFC2119].


   The layout of this document is as follows:

   - Section 1 contains an overview of PKI management;
   - Section 2 contains discussion of assumptions and restrictions;
   - Section 3 contains data structures used for PKI management messages;
   - Section 4 defines the functions that are to be carried out in PKI
     management by conforming implementations;
   - Section 5 describes a simple protocol for transporting PKI messages;
   - the Appendices specify profiles for conforming implementations and
     provide an ASN.1 module containing the syntax for all messages
     defined in this specification.

Top       Page 2 
1 PKI Management Overview

   The PKI must be structured to be consistent with the types of
   individuals who must administer it.  Providing such administrators
   with unbounded choices not only complicates the software required but
   also increases the chances that a subtle mistake by an administrator
   or software developer will result in broader compromise. Similarly,
   restricting administrators with cumbersome mechanisms will cause them
   not to use the PKI.

   Management protocols are REQUIRED to support on-line interactions
   between Public Key Infrastructure (PKI) components.  For example, a
   management protocol might be used between a Certification Authority
   (CA) and a client system with which a key pair is associated, or
   between two CAs that issue cross-certificates for each other.

1.1 PKI Management Model

   Before specifying particular message formats and procedures we first
   define the entities involved in PKI management and their interactions
   (in terms of the PKI management functions required).  We then group
   these functions in order to accommodate different identifiable types
   of end entities.

1.2 Definitions of PKI Entities

   The entities involved in PKI management include the end entity (i.e.,
   the entity to be named in the subject field of a certificate) and the
   certification authority (i.e., the entity named in the issuer field
   of a certificate). A registration authority MAY also be involved in
   PKI management.

1.2.1 Subjects and End Entities

   The term "subject" is used here to refer to the entity named in the
   subject field of a certificate; when we wish to distinguish the tools
   and/or software used by the subject (e.g., a local certificate
   management module) we will use the term "subject equipment". In
   general, the term "end entity" (EE) rather than subject is preferred
   in order to avoid confusion with the field name.

   It is important to note that the end entities here will include not
   only human users of applications, but also applications themselves
   (e.g., for IP security). This factor influences the protocols which
   the PKI management operations use; for example, application software
   is far more likely to know exactly which certificate extensions are
   required than are human users. PKI management entities are also end
   entities in the sense that they are sometimes named in the subject

Top       Page 3 
   field of a certificate or cross-certificate. Where appropriate, the
   term "end-entity" will be used to refer to end entities who are not
   PKI management entities.

   All end entities require secure local access to some information --
   at a minimum, their own name and private key, the name of a CA which
   is directly trusted by this entity and that CA's public key (or a
   fingerprint of the public key where a self-certified version is
   available elsewhere). Implementations MAY use secure local storage
   for more than this minimum (e.g., the end entity's own certificate or
   application-specific information). The form of storage will also vary
   -- from files to tamper-resistant cryptographic tokens.  Such local
   trusted storage is referred to here as the end entity's Personal
   Security Environment (PSE).

   Though PSE formats are beyond the scope of this document (they are
   very dependent on equipment, et cetera), a generic interchange format
   for PSEs is defined here - a certification response message MAY be

1.2.2 Certification Authority

   The certification authority (CA) may or may not actually be a real
   "third party" from the end entity's point of view. Quite often, the
   CA will actually belong to the same organization as the end entities
   it supports.

   Again, we use the term CA to refer to the entity named in the issuer
   field of a certificate; when it is necessary to distinguish the
   software or hardware tools used by the CA we use the term "CA

   The CA equipment will often include both an "off-line" component and
   an "on-line" component, with the CA private key only available to the
   "off-line" component. This is, however, a matter for implementers
   (though it is also relevant as a policy issue).

   We use the term "root CA" to indicate a CA that is directly trusted
   by an end entity; that is, securely acquiring the value of a root CA
   public key requires some out-of-band step(s). This term is not meant
   to imply that a root CA is necessarily at the top of any hierarchy,
   simply that the CA in question is trusted directly.

   A "subordinate CA" is one that is not a root CA for the end entity in
   question. Often, a subordinate CA will not be a root CA for any
   entity but this is not mandatory.

Top       Page 4 
1.2.3 Registration Authority

   In addition to end-entities and CAs, many environments call for the
   existence of a Registration Authority (RA) separate from the
   Certification Authority. The functions which the registration
   authority may carry out will vary from case to case but MAY include
   personal authentication, token distribution, revocation reporting,
   name assignment, key generation, archival of key pairs, et cetera.

   This document views the RA as an OPTIONAL component - when it is not
   present the CA is assumed to be able to carry out the RA's functions
   so that the PKI management protocols are the same from the end-
   entity's point of view.

   Again, we distinguish, where necessary, between the RA and the tools
   used (the "RA equipment").

   Note that an RA is itself an end entity. We further assume that all
   RAs are in fact certified end entities and that RAs have private keys
   that are usable for signing. How a particular CA equipment identifies
   some end entities as RAs is an implementation issue (i.e., this
   document specifies no special RA certification operation). We do not
   mandate that the RA is certified by the CA with which it is
   interacting at the moment (so one RA may work with more than one CA
   whilst only being certified once).

   In some circumstances end entities will communicate directly with a
   CA even where an RA is present. For example, for initial registration
   and/or certification the subject may use its RA, but communicate
   directly with the CA in order to refresh its certificate.

1.3 PKI Management Requirements

   The protocols given here meet the following requirements on PKI

      1. PKI management must conform to the ISO 9594-8 standard and the
         associated amendments (certificate extensions)

      2. PKI management must conform to the other parts of this series.

      3. It must be possible to regularly update any key pair without
         affecting any other key pair.

      4. The use of confidentiality in PKI management protocols must be
         kept to a minimum in order to ease regulatory problems.

Top       Page 5 
      5. PKI management protocols must allow the use of different
         industry-standard cryptographic algorithms, (specifically
         including RSA, DSA, MD5, SHA-1) -- this means that any given
         CA, RA, or end entity may, in principle, use whichever
         algorithms suit it for its own key pair(s).

      6. PKI management protocols must not preclude the generation of
         key pairs by the end-entity concerned, by an RA, or by a CA --
         key generation may also occur elsewhere, but for the purposes
         of PKI management we can regard key generation as occurring
         wherever the key is first present at an end entity, RA, or CA.

      7. PKI management protocols must support the publication of
         certificates by the end-entity concerned, by an RA, or by a CA.
         Different implementations and different environments may choose
         any of the above approaches.

      8. PKI management protocols must support the production of
         Certificate Revocation Lists (CRLs) by allowing certified end
         entities to make requests for the revocation of certificates -
         this must be done in such a way that the denial-of-service
         attacks which are possible are not made simpler.

      9. PKI management protocols must be usable over a variety of
         "transport" mechanisms, specifically including mail, http,
         TCP/IP and ftp.

      10. Final authority for certification creation rests with the CA;
          no RA or end-entity equipment can assume that any certificate
          issued by a CA will contain what was requested -- a CA may
          alter certificate field values or may add, delete or alter
          extensions according to its operating policy. In other words,
          all PKI entities (end-entities, RAs, and CAs) must be capable
          of handling responses to requests for certificates in which
          the actual certificate issued is different from that requested
          (for example, a CA may shorten the validity period requested).
          Note that policy may dictate that the CA must not publish or
          otherwise distribute the certificate until the requesting
          entity has reviewed and accepted the newly-created certificate
          (typically through use of the PKIConfirm message).

      11. A graceful, scheduled change-over from one non-compromised CA
          key pair to the next (CA key update) must be supported (note
          that if the CA key is compromised, re-initialization must be
          performed for all entities in the domain of that CA). An end
          entity whose PSE contains the new CA public key (following a
          CA key update) must also be able to verify certificates
          verifiable using the old public key. End entities who directly

Top       Page 6 
          trust the old CA key pair must also be able to verify
          certificates signed using the new CA private key.  (Required
          for situations where the old CA public key is "hardwired" into
          the end entity's cryptographic equipment).

      12. The Functions of an RA may, in some implementations or
          environments, be carried out by the CA itself. The protocols
          must be designed so that end entities will use the same
          protocol (but, of course, not the same key!) regardless of
          whether the communication is with an RA or CA.

      13. Where an end entity requests a certificate containing a given
          public key value, the end entity must be ready to demonstrate
          possession of the corresponding private key value. This may be
          accomplished in various ways, depending on the type of
          certification request. See Section 2.3, "Proof of Possession
          of Private Key", for details of the in-band methods defined
          for the PKIX-CMP (i.e., Certificate Management Protocol)

PKI Management Operations

   The following diagram shows the relationship between the entities
   defined above in terms of the PKI management operations. The letters
   in the diagram indicate "protocols" in the sense that a defined set
   of PKI management messages can be sent along each of the lettered

Top       Page 7 
      +---+     cert. publish        +------------+      j
      |   |  <---------------------  | End Entity | <-------
      | C |             g            +------------+      "out-of-band"
      |   |                            | ^                loading
      | e |                            | |      initial
      | r |                          a | | b     registration/
      | t |                            | |       certification
      |   |                            | |      key pair recovery
      | / |                            | |      key pair update
      |   |                            | |      certificate update
      | C |  PKI "USERS"               V |      revocation request
      | R | -------------------+-+-----+-+------+-+-------------------
      | L |  PKI MANAGEMENT    | ^              | ^
      |   |    ENTITIES      a | | b          a | | b
      |   |                    V |              | |
      | R |             g   +------+    d       | |
      | e |   <------------ | RA   | <-----+    | |
      | p |      cert.      |      | ----+ |    | |
      | o |       publish   +------+   c | |    | |
      | s |                              | |    | |
      | i |                              V |    V |
      | t |          g                 +------------+   i
      | o |   <------------------------|     CA     |------->
      | r |          h                 +------------+  "out-of-band"
      | y |      cert. publish              | ^         publication
      |   |      CRL publish                | |
      +---+                                 | |    cross-certification
                                          e | | f  cross-certificate
                                            | |       update
                                            | |
                                            V |
                                          | CA-2 |

                           Figure 1 - PKI Entities

   At a high level the set of operations for which management messages
   are defined can be grouped as follows.

      1 CA establishment: When establishing a new CA, certain steps are
        required (e.g., production of initial CRLs, export of CA public

      2 End entity initialization: this includes importing a root CA
        public key and requesting information about the options
        supported by a PKI management entity.

Top       Page 8 
      3 Certification: various operations result in the creation of new

        3.1 initial registration/certification: This is the process
            whereby  an end entity first makes itself known to a CA or
            RA, prior to the CA issuing a certificate or certificates
            for that end entity. The end result of this process (when it
            is successful) is that a CA issues a certificate for an end
            entity's public key, and returns that certificate to the end
            entity and/or posts that certificate in a public repository.
            This process may, and typically will, involve multiple
            "steps", possibly including an initialization of the end
            entity's equipment. For example, the end entity's equipment
            must be securely initialized with the public key of a CA, to
            be used in validating certificate paths.  Furthermore, an
            end entity typically needs to be initialized with its own
            key pair(s).

        3.2 key pair update:  Every key pair needs to be updated
            regularly (i.e., replaced with a new key pair), and a new
            certificate needs to be issued.

        3.3 certificate update: As certificates expire they may be
            "refreshed" if nothing relevant in the environment has

        3.4 CA key pair update: As with end entities, CA key pairs need
            to be updated regularly; however, different mechanisms are

        3.5 cross-certification request:  One CA requests issuance of a
            cross-certificate from another CA.  For the purposes of this
            standard, the following terms are defined.  A "cross-
            certificate" is a certificate in which the subject CA and
            the issuer CA are distinct and SubjectPublicKeyInfo contains
            a verification key (i.e., the certificate has been issued
            for the subject CA's signing key pair).  When it is
            necessary to distinguish more finely, the following terms
            may be used: a cross-certificate is called an "inter-domain
            cross-certificate" if the subject and issuer CAs belong to
            different administrative domains; it is called an "intra-
            domain cross-certificate" otherwise.

Top       Page 9 

   Note 1. The above definition of "cross-certificate" aligns with the
   defined term "CA-certificate" in X.509.  Note that this term is not
   to be confused with the X.500 "cACertificate" attribute type, which
   is unrelated.

   Note 2. In many environments the term "cross-certificate", unless
   further qualified, will be understood to be synonymous with "inter-
   domain cross-certificate" as defined above.

   Note 3. Issuance of cross-certificates may be, but is not
   necessarily, mutual; that is, two CAs may issue cross-certificates
   for each other.

        3.6 cross-certificate update: Similar to a normal certificate
            update but involving a cross-certificate.

      4 Certificate/CRL discovery operations: some PKI management
        operations result in the publication of certificates or CRLs:

        4.1 certificate publication: Having gone to the trouble of
            producing a certificate, some means for publishing it is
            needed.  The "means" defined in PKIX MAY involve the
            messages specified in Sections 3.3.13 - 3.3.16, or MAY
            involve other methods (LDAP, for example) as described in
            the "Operational Protocols" documents of the PKIX series of

        4.2 CRL publication: As for certificate publication.

      5 Recovery operations: some PKI management operations are used
        when an end entity has "lost" its PSE:

        5.1 key pair recovery:  As an option, user client key materials
            (e.g., a user's private key used for decryption purposes)
            MAY be backed up by a CA, an RA, or a key backup system
            associated with a CA or RA. If an entity needs to recover
            these backed up key materials (e.g., as a result of a
            forgotten password or a lost key chain file), a  protocol
            exchange may be needed to support such recovery.

      6 Revocation operations: some PKI operations result in the
        creation of new CRL entries and/or new CRLs:

        6.1 revocation request:  An authorized person advises a CA of an
            abnormal situation requiring certificate revocation.

Top       Page 10 
      7 PSE operations: whilst the definition of PSE operations (e.g.,
        moving a PSE, changing a PIN, etc.) are beyond the scope of this
        specification, we do define a PKIMessage (CertRepMessage) which
        can form the basis of such operations.

   Note that on-line protocols are not the only way of implementing the
   above operations.  For all operations there are off-line methods of
   achieving the same result, and this specification does not mandate
   use of on-line protocols.  For example, when hardware tokens are
   used, many of the operations MAY be achieved as part of the physical
   token delivery.

   Later sections define a set of standard messages supporting the above
   operations.  The protocols for conveying these exchanges in different
   environments (file based, on-line, E-mail, and WWW) is also

2. Assumptions and restrictions

2.1 End entity initialization

   The first step for an end entity in dealing with PKI management
   entities is to request information about the PKI functions supported
   and to securely acquire a copy of the relevant root CA public key(s).

2.2 Initial registration/certification

   There are many schemes that can be used to achieve initial
   registration and certification of end entities. No one method is
   suitable for all situations due to the range of policies which a CA
   may implement and the variation in the types of end entity which can

   We can however, classify the initial registration / certification
   schemes that are supported by this specification. Note that the word
   "initial", above, is crucial - we are dealing with the situation
   where the end entity in question has had no previous contact with the
   PKI. Where the end entity already possesses certified keys then some
   simplifications/alternatives are possible.

   Having classified the schemes that are supported by this
   specification we can then specify some as mandatory and some as
   optional. The goal is that the mandatory schemes cover a sufficient
   number of the cases which will arise in real use, whilst the optional
   schemes are available for special cases which arise less frequently.
   In this way we achieve a balance between flexibility and ease of

Top       Page 11 
   We will now describe the classification of initial registration /
   certification schemes.

2.2.1 Criteria used Initiation of registration / certification

   In terms of the PKI messages which are produced we can regard the
   initiation of the initial registration / certification exchanges as
   occurring wherever the first PKI message relating to the end entity
   is produced. Note that the real-world initiation of the registration
   / certification procedure may occur elsewhere (e.g., a personnel
   department may telephone an RA operator).

   The possible locations are at the end entity, an RA, or a CA. End entity message origin authentication

   The on-line messages produced by the end entity that requires a
   certificate may be authenticated or not. The requirement here is to
   authenticate the origin of any messages from the end entity to the
   PKI (CA/RA).

   In this specification, such authentication is achieved by the PKI
   (CA/RA) issuing the end entity with a secret value (initial
   authentication key) and reference value (used to identify the
   transaction) via some out-of-band means. The initial authentication
   key can then be used to protect relevant PKI messages.

   We can thus classify the initial registration/certification scheme
   according to whether or not the on-line end entity -> PKI messages
   are authenticated or not.

   Note 1: We do not discuss the authentication of the PKI -> end entity
   messages here as this is always REQUIRED. In any case, it can be
   achieved simply once the root-CA public key has been installed at the
   end entity's equipment or it can be based on the initial
   authentication key.

   Note 2: An initial registration / certification procedure can be
   secure where the messages from the end entity are authenticated via
   some out- of-band means (e.g., a subsequent visit). Location of key generation

   In this specification, "key generation" is regarded as occurring
   wherever either the public or private component of a key pair first
   occurs in a PKIMessage. Note that this does not preclude a

Top       Page 12 
   centralized key generation service - the actual key pair MAY have
   been generated elsewhere and transported to the end entity, RA, or CA
   using a (proprietary or standardized) key generation request/response
   protocol (outside the scope of this specification).

   There are thus three possibilities for the location of "key
   generation":  the end entity, an RA, or a CA. Confirmation of successful certification

   Following the creation of an initial certificate for an end entity,
   additional assurance can be gained by having the end entity
   explicitly confirm successful receipt of the message containing (or
   indicating the creation of) the certificate. Naturally, this
   confirmation message must be protected (based on the initial
   authentication key or other means).

   This gives two further possibilities: confirmed or not.

2.2.2 Mandatory schemes

   The criteria above allow for a large number of initial registration /
   certification schemes. This specification mandates that conforming CA
   equipment, RA equipment, and EE equipment MUST support the second
   scheme listed below. Any entity MAY additionally support other
   schemes, if desired. Centralized scheme

   In terms of the classification above, this scheme is, in some ways,
   the simplest possible, where:

   - initiation occurs at the certifying CA;
   - no on-line message authentication is required;
   - "key generation" occurs at the certifying CA (see Section;
   - no confirmation message is required.

   In terms of message flow, this scheme means that the only message
   required is sent from the CA to the end entity. The message must
   contain the entire PSE for the end entity. Some out-of-band means
   must be provided to allow the end entity to authenticate the message
   received and decrypt any encrypted values.

Top       Page 13 Basic authenticated scheme

   In terms of the classification above, this scheme is where:

   - initiation occurs at the end entity;
   - message authentication is REQUIRED;
   - "key generation" occurs at the end entity (see Section;
   - a confirmation message is REQUIRED.

   In terms of message flow, the basic authenticated scheme is as

      End entity                                          RA/CA
      ==========                                      =============
           out-of-band distribution of Initial Authentication
           Key (IAK) and reference value (RA/CA -> EE)
      Key generation
      Creation of certification request
      Protect request with IAK
                    -->>--certification request-->>--
                                                     verify request
                                                     process request
                                                     create response
                    --<<--certification response--<<--
      handle response
      create confirmation
                    -->>--confirmation message-->>--
                                                     verify confirmation

   (Where verification of the confirmation message fails, the RA/CA MUST
   revoke the newly issued certificate if it has been published or
   otherwise made available.)

2.3 Proof of Possession (POP) of Private Key

   In order to prevent certain attacks and to allow a CA/RA to properly
   check the validity of the binding between an end entity and a key
   pair, the PKI management operations specified here make it possible
   for an end entity to prove that it has possession of (i.e., is able
   to use) the private key corresponding to the public key for which a
   certificate is requested.  A given CA/RA is free to choose how to
   enforce POP (e.g., out-of-band procedural means versus PKIX-CMP in-
   band messages) in its certification exchanges (i.e., this may be a
   policy issue).  However, it is REQUIRED that CAs/RAs MUST enforce POP
   by some means because there are currently many non-PKIX operational
   protocols in use (various electronic mail protocols are one example)
   that do not explicitly check the binding between the end entity and
   the private key.  Until operational protocols that do verify the

Top       Page 14 
   binding (for signature, encryption, and key agreement key pairs)
   exist, and are ubiquitous, this binding can only be assumed to have
   been verified by the CA/RA. Therefore, if the binding is not verified
   by the CA/RA, certificates in the Internet Public-Key Infrastructure
   end up being somewhat less meaningful.

   POP is accomplished in different ways depending upon the type of key
   for which a certificate is requested. If a key can be used for
   multiple purposes (e.g., an RSA key) then any appropriate method MAY
   be used (e.g., a key which may be used for signing, as well as other
   purposes, SHOULD NOT be sent to the CA/RA in order to prove

   This specification explicitly allows for cases where an end entity
   supplies the relevant proof to an RA and the RA subsequently attests
   to the CA that the required proof has been received (and validated!).
   For example, an end entity wishing to have a signing key certified
   could send the appropriate signature to the RA which then simply
   notifies the relevant CA that the end entity has supplied the
   required proof. Of course, such a situation may be disallowed by some
   policies (e.g., CAs may be the only entities permitted to verify POP
   during certification).

2.3.1 Signature Keys

   For signature keys, the end entity can sign a value to prove
   possession of the private key.

2.3.2 Encryption Keys

   For encryption keys, the end entity can provide the private key to
   the CA/RA, or can be required to decrypt a value in order to prove
   possession of the private key (see Section 3.2.8). Decrypting a value
   can be achieved either directly or indirectly.

   The direct method is for the RA/CA to issue a random challenge to
   which an immediate response by the EE is required.

   The indirect method is to issue a certificate which is encrypted for
   the end entity (and have the end entity demonstrate its ability to
   decrypt this certificate in the confirmation message). This allows a
   CA to issue a certificate in a form which can only be used by the
   intended end entity.

   This specification encourages use of the indirect method because this
   requires no extra messages to be sent (i.e., the proof can be
   demonstrated using the {request, response, confirmation} triple of

Top       Page 15 
2.3.3 Key Agreement Keys

   For key agreement keys, the end entity and the PKI management entity
   (i.e., CA or RA) must establish a shared secret key in order to prove
   that the end entity has possession of the private key.

   Note that this need not impose any restrictions on the keys that can
   be certified by a given CA -- in particular, for Diffie-Hellman keys
   the end entity may freely choose its algorithm parameters -- provided
   that the CA can generate a short-term (or one-time) key pair with the
   appropriate parameters when necessary.

2.4 Root CA key update

   This discussion only applies to CAs that are a root CA for some end

   The basis of the procedure described here is that the CA protects its
   new public key using its previous private key and vice versa. Thus
   when a CA updates its key pair it must generate two extra
   cACertificate attribute values if certificates are made available
   using an X.500 directory (for a total of four:  OldWithOld;
   OldWithNew; NewWithOld; and NewWithNew).

   When a CA changes its key pair those entities who have acquired the
   old CA public key via "out-of-band" means are most affected. It is
   these end entities who will need access to the new CA public key
   protected with the old CA private key. However, they will only
   require this for a limited period (until they have acquired the new
   CA public key via the "out-of-band" mechanism). This will typically
   be easily achieved when these end entities' certificates expire.

   The data structure used to protect the new and old CA public keys is
   a standard certificate (which may also contain extensions). There are
   no new data structures required.

   Note 1. This scheme does not make use of any of the X.509 v3
   extensions as it must be able to work even for version 1
   certificates. The presence of the KeyIdentifier extension would make
   for efficiency improvements.

   Note 2. While the scheme could be generalized to cover cases where
   the CA updates its key pair more than once during the validity period
   of one of its end entities' certificates, this generalization seems
   of dubious value. Not having this generalization simply means that
   the validity period of a CA key pair must be greater than the
   validity period of any certificate issued by that CA using that key

Top       Page 16 
   Note 3.This scheme forces end entities to acquire the new CA public
   key on the expiry of the last certificate they owned that was signed
   with the old CA private key (via the "out-of-band" means).
   Certificate and/or key update operations occurring at other times do
   not necessarily require this (depending on the end entity's

2.4.1 CA Operator actions

   To change the key of the CA, the CA operator does the following:

      1. Generate a new key pair;

      2. Create a certificate containing the old CA public key signed
         with the new private key (the "old with new" certificate);

      3. Create a certificate containing the new CA public key signed
         with the old private key (the "new with old" certificate);

      4. Create a certificate containing the new CA public key signed
         with the new private key (the "new with new" certificate);

      5. Publish these new certificates via the directory and/or other
         means (perhaps using a CAKeyUpdAnn message);

      6. Export the new CA public key so that end entities may acquire
         it using the "out-of-band" mechanism (if required).

   The old CA private key is then no longer required. The old CA public
   key will however remain in use for some time. The time when the old
   CA public key is no longer required (other than for non-repudiation)
   will be when all end entities of this CA have securely acquired the
   new CA public key.

   The "old with new" certificate must have a validity period starting
   at the generation time of the old key pair and ending at the expiry
   date of the old public key.

   The "new with old" certificate must have a validity period starting
   at the generation time of the new key pair and ending at the time by
   which all end entities of this CA will securely possess the new CA
   public key (at the latest, the expiry date of the old public key).

   The "new with new" certificate must have a validity period starting
   at the generation time of the new key pair and ending at the time by
   which the CA will next update its key pair.

Top       Page 17 
2.4.2 Verifying Certificates.

   Normally when verifying a signature, the verifier verifies (among
   other things) the certificate containing the public key of the
   signer. However, once a CA is allowed to update its key there are a
   range of new possibilities. These are shown in the table below.

               Repository contains NEW     Repository contains only OLD
                 and OLD public keys        public key (due to, e.g.,
                                             delay in publication)

                  PSE      PSE Contains  PSE Contains    PSE Contains
               Contains     OLD public    NEW public      OLD public
              NEW public       key            key            key

   Signer's   Case 1:      Case 3:       Case 5:        Case 7:
   certifi-   This is      In this case  Although the   In this case
   cate is    the          the verifier  CA operator    the CA
   protected  standard     must access   has not        operator  has
   using NEW  case where   the           updated the    not updated
   public     the          directory in  directory the  the directory
   key        verifier     order to get  verifier can   and so the
              can          the value of  verify the     verification
              directly     the NEW       certificate    will FAIL
              verify the   public key    directly -
              certificate                this is thus
              without                    the same as
              using the                  case 1.

   Signer's   Case 2:      Case 4:       Case 6:        Case 8:
   certifi-   In this      In this case  The verifier   Although the
   cate is    case the     the verifier  thinks this    CA operator
   protected  verifier     can directly  is the         has not
   using OLD  must         verify the    situation of   updated the
   public     access the   certificate   case 2 and     directory the
   key        directory    without       will access    verifier can
              in order     using the     the            verify the
              to get the   directory     directory;     certificate
              value of                   however, the   directly -
              the OLD                    verification   this is thus
              public key                 will FAIL      the same as
                                                        case 4.

Top       Page 18 Verification in cases 1, 4, 5 and 8.

   In these cases the verifier has a local copy of the CA public key
   which can be used to verify the certificate directly. This is the
   same as the situation where no key change has occurred.

   Note that case 8 may arise between the time when the CA operator has
   generated the new key pair and the time when the CA operator stores
   the updated attributes in the directory. Case 5 can only arise if the
   CA operator has issued both the signer's and verifier's certificates
   during this "gap" (the CA operator SHOULD avoid this as it leads to
   the failure cases described below). Verification in case 2.

   In case 2 the verifier must get access to the old public key of the
   CA. The verifier does the following:

      1. Look up the caCertificate attribute in the directory and pick
         the OldWithNew certificate (determined based on validity
      2. Verify that this is correct using the new CA key (which the
         verifier has locally);
      3. If correct, check the signer's certificate using the old CA

   Case 2 will arise when the CA operator has issued the signer's
   certificate, then changed key and then issued the verifier's
   certificate, so it is quite a typical case. Verification in case 3.

   In case 3 the verifier must get access to the new public key of the
   CA. The verifier does the following:

      1. Look up the CACertificate attribute in the directory and pick
         the NewWithOld certificate (determined based on validity
      2. Verify that this is correct using the old CA key (which the
         verifier has stored locally);
      3. If correct, check the signer's certificate using the new CA

   Case 3 will arise when the CA operator has issued the verifier's
   certificate, then changed key and then issued the signer's
   certificate, so it is also quite a typical case.

Top       Page 19 Failure of verification in case 6.

   In this case the CA has issued the verifier's PSE containing the new
   key without updating the directory attributes. This means that the
   verifier has no means to get a trustworthy version of the CA's old
   key and so verification fails.

   Note that the failure is the CA operator's fault. Failure of verification in case 7.

   In this case the CA has issued the signer's certificate protected
   with the new key without updating the directory attributes. This
   means that the verifier has no means to get a trustworthy version of
   the CA's new key and so verification fails.

   Note that the failure is again the CA operator's fault.

2.4.3 Revocation - Change of CA key

   As we saw above the verification of a certificate becomes more
   complex once the CA is allowed to change its key. This is also true
   for revocation checks as the CA may have signed the CRL using a newer
   private key than the one that is within the user's PSE.

   The analysis of the alternatives is as for certificate verification.

(page 19 continued on part 2)

Next RFC Part