Network Working Group                                    S. Farrell, Ed.
Request for Comments: 3767                        Trinity College Dublin
Category: Standards Track                                      June 2004


               Securely Available Credentials Protocol

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 (2004).

Abstract

  This document describes a protocol whereby a user can acquire
  cryptographic credentials (e.g., private keys, PKCS #15 structures)
  from a credential server, using a workstation that has locally
  trusted software installed, but with no user-specific configuration.
  The protocol's payloads are described in XML.  This memo also
  specifies a Blocks Extensible Exchange Protocol (BEEP) profile of the
  protocol.  Security requirements are  met by mandating support for
  TLS and/or DIGEST-MD5 (through BEEP).

Table Of Contents

  1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  2
  2.  The Protocol. . . . . .  . . . . . . . . . . . . . . . . . . .  3
  3.  BEEP Profile for SACRED. . . . . . . . . . . . . . . . . . . .  9
  4.  IANA Considerations. . . . . . . . . . . . . . . . . . . . . . 12
  5.  Security Considerations. . . . . . . . . . . . . . . . . . . . 13
  6.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 15
  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . 16
  Appendix A: XML Schema . . . . . . . . . . . . . . . . . . . . . . 17
  Appendix B: An Example of Tuning with BEEP . . . . . . . . . . . . 20
  Appendix C: Provision SACRED using other Protocols . . . . . . . . 23
  Editor's Address . . . . . . . . . . . . . . . . . . . . . . . . . 24
  Full Copyright Statement. . . . . . . . . . . .  . . . . . . . . . 25







Farrell                     Standards Track                     [Page 1]

RFC 3767              Secure Credentials Protocol              June 2004


1.  Introduction

  Digital credentials, such as private keys and corresponding
  certificates, are used to support various Internet protocols, e.g.
  S/MIME, IPSec, and TLS.  In a number of environments, end users wish
  to use the same credentials on different end-user devices.  In a
  "typical" desktop environment, the user already has many tools
  available to allow import/export of these credentials.  However, this
  is not very practical.  In addition, with some devices, especially
  wireless and other more constrained devices, the tools required
  simply do not exist.

  This document describes a protocol for the secure exchange of such
  credentials and is a realization of the abstract protocol framework
  described in [RFC3760].

  Many user-chosen passwords are vulnerable to dictionary attacks.  So
  the SACRED protocol is designed to give no information with which an
  attacker can acquire information for launching a dictionary attack,
  whether by eavesdropping or by impersonating either the client or
  server.

  The protocol also allows a user to create or delete an account,
  change her account password and/or credentials, and upload the new
  values to the server.  The protocol ensures that only someone that
  knew the old account password is able to modify the credentials as
  stored on the credential server.  The protocol does not preclude
  configuring a server to disallow some operations (e.g. credential
  upload) for some users.  The account management operations as a whole
  are optional implementations for both credential servers and clients.

  Note that there are potentially two "passwords" involved when using
  this protocol - the first used to authenticate the user to the
  credential server, and the second to decrypt (parts of) the
  credential following a download operation.  Where the context
  requires it, we refer to the former as the account password and the
  latter as the credential password.

  Using a protocol such as this is somewhat less secure than using a
  smart card, but can be used until smart cards and smart card readers
  on workstations become ubiquitous, and can be useful even after smart
  cards are ubiquitous, as a backup strategy when a user's smart card
  is lost or malfunctioning.

  The protocol sets out to meet the requirements in [REQS].
  Cryptographic credentials may take the form of private keys, PKCS #15
  [PKCS15], or structures.  As stated, a profile based on BEEP [BEEP]
  is specified for message transport and security (integrity,



Farrell                     Standards Track                     [Page 2]

RFC 3767              Secure Credentials Protocol              June 2004


  authentication, and confidentiality).  In that case, the security
  requirements are met by mandating support (via BEEP) for TLS [TLS]
  and/or DIGEST-MD5 [DIGEST-MD5].

  We assume the only authentication information available to the user
  is a username and password.

  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 [RFC2119].

2.  The Protocol

  This section defines the account management and "run-time" operations
  for the SACRED protocol.

  It also describes the message formats used, which are described in
  XML [XMLSCHEMA].  Appendix A provides an XML schema for these
  elements.

  The approach taken here is to define SACRED elements that are
  compatible with the elements used in [XKMS] and [XMLDSIG], so that an
  implementation of this protocol can easily also support XKMS, and
  vice versa.

  It is also intended that other SACRED protocol instances (e.g. using
  a different authentication scheme, credential format, or transport
  protocol) could re-use many of the definitions here.

2.1.  Account Management Operations

  These operations MAY be implemented, that is, they are OPTIONAL.

2.1.1.  Information Request

  This operation does NOT REQUIRE authentication.

  The purpose of this operation is to provide the client with the
  values required for account creation.

  The client sends an InfoRequest message (which has no content).

  The server responds with an InfoResponse message which contains the
  authentication mechanism parameters for the server and the list of
  supported ProcessInfo types.  For DIGEST-MD5, this consists of the
  list of realms (each as an XML element named "Realm") which the
  server supports.  There MUST be at least one realm specified.




Farrell                     Standards Track                     [Page 3]

RFC 3767              Secure Credentials Protocol              June 2004


  Clients MUST be able to select one from a list of Realms and MUST be
  able to disregard any other information present (allowed for
  extensibility).

2.1.2.  Create Account

  This operation REQUIRES server authentication.

  The purpose of this operation is to setup a new account on the
  server.  The information required for a "new" account will depend on
  the SASL [SASL] mechanism used.

  The client sends a CreateAccountRequest, which contains the account
  name (e.g. username).  It also contains the elements required to
  create an account for a particular authentication mechanism.  The
  actual information is defined according to the authentication
  mechanism.  For DIGEST-MD5, this consists of the password verifier
  (the hashed username, password and realm) and the chosen realm.
  Although more than one set of such data is allowed by the data
  structures defined in the appendix, clients SHOULD only include one
  here.

  The server responds with an error or acknowledgement message.

2.1.3.  Remove Account

  This operation REQUIRES mutual authentication.

  The purpose of this operation is to delete the entire account.

  The client sends a RemoveAccountRequest message (which has no
  content) to the server.

  The server MUST delete all information relating to the account and
  respond with an error or acknowledgement message.

2.1.4.  Modify Account

  This operation REQUIRES mutual authentication.

  The purpose of this operation is to allow the client to change the
  information required for authentication.  The information required
  will depend on the authentication method used.








Farrell                     Standards Track                     [Page 4]

RFC 3767              Secure Credentials Protocol              June 2004


  The client sends a ModifyAccountRequest message, which contains the
  elements required to change the authentication information for the
  account, for a particular authentication mechanism.  The actual
  information is defined according to the authentication mechanism. For
  [DIGEST-MD5], it will consist of a realm and password verifier value.

  Once the account information has been changed, the server will
  respond with an error or acknowledgement message.

2.2.  "Run-time" Operations

  These operations MUST be supported by all conformant implementations.

2.2.1.  Credential Upload

  This operation REQUIRES mutual authentication.

  The purpose of this operation is to allow the client to deposit a
  credential with the server.

  The client sends an UploadRequest message to the server which MUST
  contain one Credential.

  If a credential with the same credential selector field as in the
  UploadRequest (a "matching" credential) already exists for the
  account, then that credential is replaced with the new credential
  from the UploadRequest.  Otherwise a "new" credential is associated
  with that account.  If a new credential is being uploaded, then the
  client SHOULD include (in LastModified) its local concept of the time
  (if it has one), or an indicator that it has no clock.  The actual
  value of LastModified can be anything, (but the element has to be
  present) since this will be overwritten by the server in any case.

  If any change is made to the stored credentials associated with the
  account, then the server MUST update the corresponding LastModified
  value (returned in DownloadResponse messages) to the current time (at
  the server).














Farrell                     Standards Track                     [Page 5]

RFC 3767              Secure Credentials Protocol              June 2004


  The LastModified value in the UploadRequest MUST be the value which
  was most recently received in a corresponding DownloadResponse for
  that credential.  This means the clients are strongly RECOMMENDED to
  only produce an UploadRequest based on recently downloaded
  credentials, since otherwise the LastModified value may be out of
  date.

  The LastModified value can also be of use in detecting conflicts.
  For example, download to platform A, download to platform B, update
  from B, update from A.  The server could detect a conflict on the
  second upload.  In this case the server MUST respond with a BEEP
  error (which SHOULD be StaleCredential).

  The server replaces the provided LastModified value with the current
  time at the server before storing the credential.  (Note that this
  means that it would be unwise for a client to include the
  LastModified field in a ClientInfo digital signature which is
  calculated over the CredentialType.)

  The server responds with an error or acknowledgement message.

2.2.2.  Credential Download

  This operation REQUIRES mutual authentication.

  The purpose of this operation is to allow a client to get one or more
  credentials from a server (the purpose of the entire protocol
  really!).

  The client sends a DownloadRequest message to the server which MAY
  contain a credential selector string for the credential.  No, or an
  empty credential selector means the request is for all credentials
  associated with the account.

  The server responds with a DownloadResponse or an error message.  A
  DownloadResponse contains one or more credential payloads, including
  the LastModified time which represents the time (at the server) when
  the last change was made to each credential associated with the
  account (e.g. subsequent to an UploadRequest).

2.2.3.  Credential Delete

  This operation REQUIRES mutual authentication.

  The purpose of this operation is to allow the client to delete one or
  all credentials associated with the account.





Farrell                     Standards Track                     [Page 6]

RFC 3767              Secure Credentials Protocol              June 2004


  The client sends a DeleteRequest message to the server which can
  contain either a CredentialSelector or an All element.

  If the DeleteRequest contains an All element, then all of the
  credentials associated with that account are deleted.

  If the DeleteRequest contains a CredentialSelector, then the request
  MAY include a LastModified value.  If the LastModified value is
  present in the DeleteRequest, then it MUST be the value which was
  most recently received in a corresponding DownloadResponse for that
  credential.  If the value does not match, then the server MUST NOT
  delete the credentials.

  If no "matching" credential exists, the server returns an error.

  The server responds to this request with an error or acknowledgement
  message.

2.3.  Miscellaneous

2.3.1.  Session Security

  Six SACRED operations are defined above.  In this section we specify
  the requirements for security for each of the operations (where
  supported).

       Operation                 Security REQUIRED
       ---------                 -----------------
       Information request       NONE
       Create account            Server authentication,
                                 Confidentiality, Integrity
       Remove account            Mutual authentication,
                                 Confidentiality, Integrity
       Modify account            Mutual authentication,
                                 Confidentiality, Integrity
       Credential upload         Mutual authentication,
                                 Confidentiality, Integrity
       Credential download       Mutual authentication,
                                 Confidentiality, Integrity
       Credential delete         Mutual authentication,
                                 Confidentiality, Integrity

  The security requirements can be met by several mechanisms.  This
  document REQUIRES credential servers to support TLS and DIGEST-MD5.
  Clients MUST support DIGEST-MD5 and TLS with server authentication.






Farrell                     Standards Track                     [Page 7]

RFC 3767              Secure Credentials Protocol              June 2004


  The mandatory-to-implement TLS cipher suite for SACRED is
  TLS_RSA_WITH_3DES-EDE_CBC_SHA.  Implementations SHOULD also support
  TLS_RSA_WITH_AES_128_CBC_SHA [TLSAES].

  When performing mutual authentication using DIGEST-MD5 for the
  client, DIGEST-MD5 MUST only be used "within" a TLS server-
  authenticated "pipe", and MUST only be used for client
  authentication.  That is, we do not use the DIGEST-MD5 security
  services (confidentiality, integrity etc.).

2.3.2.  Handling Multiple Credentials for an Account

  When more than one credential is stored under a single account, the
  client can select a single credential using the optional credential
  selector string.

  There is no concept of a "default credential" - all credentials MUST
  have an associated selector unique for that account.  The selector is
  REQUIRED for upload requests and OPTIONAL for download requests.  If
  the selector is omitted in a download request, it MUST be interpreted
  as a request for all the stored credentials.

  An empty selector string value (i.e. "") in a credential download
  request is to be interpreted as if the selector string were omitted,
  i.e. a download request containing this is a request for all
  credentials.

  It is an error to have more than one credential stored under the same
  account where both have the same credential selector string.

2.3.3.  Common Fields

  All messages sent to the server MAY contain ProcessInfo values.  This
  field MAY be used by other specifications or for vendor extensions.
  For example, a server might require clients to include a phone number
  in this field.  The information response message contains a list of
  the types of ProcessInfo that the server supports.  This
  extensibility scheme is similar to that used in [XKMS] and [XBULK].

  Where no specific response message is defined for an operation (e.g.
  for UploadRequest), then the transport will indicate success or
  failure.

  All of the response messages defined here MAY contain a Status
  string, containing a value intended for human consumption.






Farrell                     Standards Track                     [Page 8]

RFC 3767              Secure Credentials Protocol              June 2004


2.3.4.  Credential Format

  A number of messages involve the Credential element.  It has the
  following fields (all optional fields may occur exactly zero or one
  times unless otherwise stated):

  -  CredentialSelector contains a string by which this particular
     credential (for this account) can be identified.
  -  PayLoad contains either a ds:KeyInfo or some other form of
     credential.  Implementations MUST support the PKCS #15 form of
     ds:KeyInfo defined below (the SacredPKCS15 element).
  -  LastModified is a string containing the time (at the server) at
     which this credential was last modified.
  -  TimeToLive (optional) is a hint clients SHOULD honor, which
     specifies the number of seconds the downloaded credential is to be
     usable.
  -  ProcessInfo (optional) MAY contain any (typed) information that
     the server is intended to process.  If the server doesn't support
     any of the ProcessInfo data, it MAY ignore that data.
  -  ClientInfo (optional) MAY contain any (typed) information that the
     client is intended to process, but which the server MUST ignore.
     If the client doesn't support any of the ClientInfo data, it MAY
     ignore that data (e.g. if the ClientInfo is device specific).

3.  BEEP Profile for SACRED

  The protocol described in this memo is realized as a [BEEP] profile.

  Future memos may define alternative versions of the BEEP profile for
  SACRED.  When a BEEP peer sends its greeting, it indicates which
  profiles it is willing to support.  Accordingly, when the BEEP client
  asks to start a channel, it indicates the versions it supports, and
  if any of these are acceptable to the BEEP server; the latter
  specifies which profile it is starting.

  Profile Identification: http://iana.org/beep/sacred

  Messages Exchanged during Channel Creation:
       InfoRequest,
       CreateAccountRequest,
       RemoveAccountRequest,
       ModifyAccountRequest,
       DownloadRequest,
       UploadRequest,
       DeleteRequest,
       InfoResponse,





Farrell                     Standards Track                     [Page 9]

RFC 3767              Secure Credentials Protocol              June 2004


       DownloadResponse,
       error,
       ok

  Messages starting one-to-one exchanges:
       InfoRequest,
       CreateAccountRequest,
       RemoveAccountRequest,
       ModifyAccountRequest,
       DownloadRequest,
       UploadRequest,
       DeleteRequest

  Messages in positive replies:
       ok,
       InfoResponse,
       DownloadResponse

  Messages in negative replies: error

  Messages in one-to-many changes: none

  Message Syntax: c.f.,Section 3

  Message Semantics: c.f., Section 2

  Contact Information: c.f., the editor's address section of this memo

3.1.  Profile Initialization

  Because all but one of the operations of the SACRED profile have
  security requirements (cf., Section 2.3.1), before starting the
  SACRED profile, the BEEP session will likely be tuned using either

         http://iana.org/beep/TLS

         or

         http://iana.org/beep/TLS followed by
         http://iana.org/SASL/DIGEST-MD5

  Appendix B gives an example of tuning a BEEP session using DIGEST-
  MD5 (i.e. it shows how to turn on BEEP security).

  Regardless, upon completion of the negotiation process, a tuning
  reset occurs in which both BEEP peers issue a new greeting.  Consult
  Section 3 of [BEEP] for an example of how a BEEP peer may choose to
  issue different greetings based on whether confidentiality is in use.



Farrell                     Standards Track                    [Page 10]

RFC 3767              Secure Credentials Protocol              June 2004


  Any of the messages listed in section 3.2 below may be exchanged
  during channel initialization (c.f., Section 2.3.1.2 of [BEEP]),
  e.g.,

       C: <start number='1'>
       C:   <profile uri='http://iana.org/beep/sacred'>
       C:             <![CDATA[<DownloadRequest ...>]]>
       C:     </profile>
       C: </start>

       S: <profile uri='http://iana.org/beep/sacred'>
       S:   <![CDATA[<DownloadResponse ...>]]>
       S: </profile>

  Note that BEEP imposes both encoding and length limitations on the
  messages that are piggybacked during channel initialization.

3.2.  Profile Exchange

  All messages are exchanged as "application/beep+xml" (c.f., Section
  6.4 of [BEEP]):

  Role         MSG                   RPY                     ERR
  ----         ---                   ---                     ---
  I            InfoRequest           InfoResponse            error
  I            CreateAccountRequest  ok                      error
  I            RemoveAccountRequest  ok                      error
  I            ModifyAccountRequest  ok                      error
  I            DownloadRequest       DownloadResponse        error
  I            UploadRequest         ok                      error
  I            DeleteRequest         Ok                      error


3.3.  Error Handling

  The "error" message from Section 2.3.1.5 of [BEEP] is used to convey
  error information.  Typically, after flagging an error, a peer will
  initiate a graceful release of the BEEP session.

  The following BEEP error reply codes from [BEEP] are to be used:

   code  Meaning
   ====  =======
   421   service not available
   450   requested action not taken (e.g., lock already in
          use)
   451   requested action aborted (e.g., local error in
          processing)



Farrell                     Standards Track                    [Page 11]

RFC 3767              Secure Credentials Protocol              June 2004


   454   temporary authentication failure
   500   general syntax error (e.g., poorly-formed XML)
   501   syntax error in parameters (e.g., non-valid XML)
   504   parameter not implemented
   530   authentication required
   534   authentication mechanism insufficient (e.g., too
          weak, sequence exhausted, etc.)
   535   authentication failure
   537   action not authorized for user
   538   authentication mechanism requires encryption
   550   requested action not taken (e.g., no requested
          profiles are acceptable)
   553   parameter invalid
   554   transaction failed (e.g., policy violation)

  The following SACRED-specific error reply codes can also be used:

   code  Meaning
   ====  =======
   555   Extension (ProcessInfo) used not supported
   556   Required extension (ProcessInfo) not present
   557   StaleCredential (A bad LastModified value was
          contained in an UploadRequest.)

3.4.  SASL Authorization Identity

  The use of the SASL authorization identity in this protocol is
  implementation-specific.  If used, the authorization identity is not
  a substitute for the credential selector field, but may be used to
  affect authorization for access to credentials.

4.  IANA Considerations

  The IANA has registered the BEEP profile specified in Section 4.

     http://iana.org/beep/sacred

  The sacred protocol SHOULD be run over port 1118.

  The GSSAPI service name (required when using SASL) for this protocol
  SHALL be "sacred".










Farrell                     Standards Track                    [Page 12]

RFC 3767              Secure Credentials Protocol              June 2004


5.  Security Considerations

  [REQS] calls for specifications to state how they address the
  vulnerabilities listed below.

     V1.   A passive attacker can watch all packets on the network and
           later carry out a dictionary attack.
           - The use of DIGEST-MD5 and/or TLS counters this
           vulnerability.
     V2.   An attacker can attempt to masquerade as a credential server
           in an attempt to get a client to reveal information online
           that allows for a later dictionary attack.
           - The use of server or mutual authentication counters this
           vulnerability.
     V3.   An attacker can attempt to get a client to decrypt a chosen
           "ciphertext" and get the client to make use of the resulting
           plaintext - the attacker may then be able to carry out a
           dictionary attack (e.g. if the plaintext resulting from
           "decryption" of a random string is used as a DSA private
           key).
           - The use of server or mutual authentication counters this
           vulnerability.
     V4.   An attacker could overwrite a repository entry so that when
           a user subsequently uses what they think is a good
           credential, they expose information about their password
           (and hence the "real" credential).
           - Server implementations SHOULD take measures to protect the
           database.  Clients MAY use the ClientInfo field to store
           e.g. a signature over the Credential, which they then verify
           before using the private component.
     V5.   An attacker can copy a credential server's repository and
           carry out a dictionary attack.
           - Server implementations SHOULD take measures to protect the
           database.
     V6.   An attacker can attempt to masquerade as a client in an
           attempt to get a server to reveal information that allows
           for a later dictionary attack.
           - The mutual authentication requirements of this protocol
           counter this to a great extent.  Additionally, credential
           servers MAY choose to provide mechanisms that protect
           against online dictionary attacks against user account
           passwords, either by repeated access attempts to a single
           user account (varying the password) or by attempting to
           access many user accounts using the same password.
     V7.   An attacker can persuade a server that a successful login
           has occurred, even if it hasn't.
           - Client authentication prevents this.




Farrell                     Standards Track                    [Page 13]

RFC 3767              Secure Credentials Protocol              June 2004


     V8.   (Upload) An attacker can overwrite someone else's
           credentials on the server.
           - Only if they know the account password already (thanks to
           mutual authentication).
     V9.   (When using password-based authentication) An attacker can
           force a password change to a known (or "weak") password.
           - Client authentication counters this.
     V10.  An attacker can attempt a man-in-the-middle attack for lots
           of reasons...
           - Mutual authentication and the encryption of subsequent
           messages prevents this.
     V11.  User enters password instead of name.
           - Since the DIGEST-MD5 mechanism is only used after TLS
           tuning, the user's name is also protected.
     V12.  An attacker could attempt various denial-of-service attacks.
           - No specific countermeasures against DoS are proposed.

  If the CreateAccountRequest message were sent over a cleartext
  channel (or otherwise exposed), then an attacker could mount a
  dictionary attack and recover the account password.  This is why the
  server authenticated TLS transport is REQUIRED for this operation.

  If someone steals the server database they can launch a dictionary
  attack.  If the dictionary attack is successful, the attacker can
  decrypt the user's credentials.  An attacker that has learned the
  user's account password can also upload new credentials, assuming the
  user is authorized to modify the credentials, because someone who
  knows the user's account password is assumed to be the user.
  However, if someone steals the server database and is unsuccessful at
  obtaining the user's account password through a dictionary attack,
  they will be unable to upload new credentials.

  Credential servers SHOULD incorporate measures that act to counter
  denial of service attacks.  In particular, they SHOULD drop inactive
  connections and minimize the use of resources by un-authenticated
  connections.  A number of recommendations are listed at [DDOS].

  Various operations in the SACRED protocol depend upon server
  authentication being provided by server authenticated TLS.  SACRED
  clients SHOULD take care that the correct server is at the far end of
  the TLS "pipe" by performing the checks which are listed in section
  3.1 of RFC 2818 [RFC2818].  Clients SHOULD also include the optional
  BEEP serverName field in their "start" message and SHOULD then ensure
  that the BEEP serverName is consistent with the checks on the TLS
  server described in RFC 2818.  Failure to carry out these checks
  could allow a spoof server access to a user's credential.





Farrell                     Standards Track                    [Page 14]

RFC 3767              Secure Credentials Protocol              June 2004


  If the SACRED account password were to be used in some other, less
  secure protocol, using DIGEST-MD5, then it might appear to be the
  case that a man-in-the-middle (MITM) attack could be mounted.
  However, this is not the case since the DIGEST-MD5 client hash
  includes a client-selected "digest-uri-value", which in SACRED's case
  will be "sacred/<serverName>".  In a MITM attack, those values will
  be something else.  A MITM attack as described is therefore thwarted,
  because digest-uri-value wouldn't match what the SACRED server is
  expecting.

6.  References

6.1.  Normative References

  [BEEP]       Rose, M., "The Blocks Extensible Exchange Protocol
               Core", RFC 3080, March 2001.

  [DIGEST-MD5] Leach, P. and C. Newman, "Using Digest Authentication as
               a SASL Mechanism", RFC 2831, May 2000.

  [PKCS15]     "PKCS #15 v1.1: Cryptographic Token Information Syntax
               Standard," RSA Laboratories, June 2000.

  [REQS]       Arsenault, A. and S. Farrell, "Securely Available
               Credentials - Requirements", RFC 3157, August 2001.

  [RFC2119]    Bradner, S., "Key words for use in RFCs to Indicate
               Requirement Levels", BCP 14, RFC 2119, March 1997.

  [SASL]       Myers, J., "Simple Authentication and Security Layer
               (SASL)", RFC 2222, October 1997.

  [TLS]        Dierks, T. and C. Allen, "The TLS Protocol - Version
               1.0", RFC 2246, January 1999.

  [TLSAES]     Chown, P., "Advanced Encryption Standard (AES)
               Ciphersuites for Transport Layer Security (TLS)", RFC
               3268, June 2002.

  [XMLDSIG]    Eastlake, 3rd, D., Reagle, J. and D. Solo, "(Extensible
               Mark-Up Language) XML-Signature Syntax and Processing",
               RFC 3275, March 2002.

  [XMLSCHEMA]  "XML Schema Part 1: Structures", D. Beech, M. Maloney,
               N. Mendelsohn, and H. Thompson.  W3C Recommendation, May
               2001.  Available at http://www.w3.org/TR/2001/REC-
               xmlschema-2-20010502/




Farrell                     Standards Track                    [Page 15]

RFC 3767              Secure Credentials Protocol              June 2004


6.2.  Informative References

  [DDOS]       "Recommendations for the Protection against Distributed
               Denial-of-Service Attacks in the Internet",
               http://www.iwar.org.uk/comsec/resources/dos/ddos_en.htm

  [RFC2818]    Rescorla, E., "HTTP over TLS", RFC 2818, May 2000.

  [RFC3760]    Gustafson, D., Just, M. and M. Nystrom, "Securely
               Available Credentials - Credential Server Framework,"
               RFC 3760, April 2004.

  [XKMS]       Hallam-Baker, P. (ed), "XML Key Management
               Specification", http://www.w3.org/TR/xkms2/

  [XBULK]      Hughes, M (ed), "XML Key Management Specification - Bulk
               Operation", http://www.w3.org/TR/xkms2-xbulk/

Acknowledgements

  Radia Perlman ([email protected]) and Charlie Kaufman
  ([email protected]) co-authored earlier versions of this
  document.  Michael Zolotarev ([email protected]) did much of the
  initial work, adapting an earlier version to the use of SRP (though
  SRP was subsequently dropped, much of the framework survives).
  Marshall Rose ([email protected]) helped out a lot, in
  particular, with the BEEP profile.  And the following people were
  actively involved in the mailing list discussions leading to this
  document:

       David Chizmadia,
       Dave Crocker ([email protected]),
       Lawrence Greenfield ([email protected]),
       Dale Gustafson ([email protected]),
       Mike Just ([email protected]),
       John Linn ([email protected]),
       Neal McBurnett ([email protected]),
       Keith Moore ([email protected]),
       RL "Bob" Morgan ([email protected]),
       Magnus Nystrom ([email protected]),
       Eamon O'Tuathail ([email protected]),
       Gareth Richards ([email protected])

  Of course, any and all errors remain the editor's responsibility.







Farrell                     Standards Track                    [Page 16]

RFC 3767              Secure Credentials Protocol              June 2004


Appendix A: XML Schema

  <?xml version="1.0" encoding="UTF-8"?>
     <schema
       targetNamespace="urn:sacred-2002-12-19"
       xmlns:ds="http://www.w3.org/2000/09/xmldsig#"
       xmlns:sacred="urn:sacred-2002-12-19"
       xmlns="http://www.w3.org/2001/XMLSchema">
       <import namespace="http://www.w3.org/2000/09/xmldsig#"
       schemaLocation=
       "http://www.w3.org/TR/xmldsig-core/xmldsig-core-schema.xsd"/>
       <!-- extensibility holes -->
       <complexType name="ProcessInfoType">
        <sequence maxOccurs="unbounded">
         <any namespace="##other"/>
        </sequence>
       </complexType>
       <element name="ProcessInfo" type="sacred:ProcessInfoType"/>
       <complexType name="ClientInfoType">
        <sequence maxOccurs="unbounded">
         <any namespace="##other"/>
        </sequence>
       </complexType>
       <element name="ClientInfo" type="sacred:ClientInfoType"/>
       <!-- Where to put authenentication information -->
       <complexType name="AuthInfoType">
        <choice maxOccurs="unbounded">
         <element name="DigestMD5AuthInfo">
          <complexType>
           <sequence>
            <element name="PasswordVerifier" type="base64Binary"/>
            <element name="Realm" type="string" />
           </sequence>
          </complexType>
         </element>
         <any namespace="##other"/>
        </choice>
       </complexType>
       <element name="AuthInfo" type="sacred:AuthInfoType"/>
       <!-- authentication mechanism parameters -->
       <complexType name="AuthParamsType">
        <choice maxOccurs="unbounded">
         <element name=" DigestMD5AuthParams">
          <complexType>
           <sequence>
            <element name="Realm" type="string"
              minOccurs="1" maxOccurs="unbounded"/>
           </sequence>



Farrell                     Standards Track                    [Page 17]

RFC 3767              Secure Credentials Protocol              June 2004


          </complexType>
         </element>
         <any namespace="##other"/>
        </choice>
       </complexType>
       <element name="AuthParams" type="sacred:AuthParamsType"/>
       <!-- Protocol messsages -->
       <!-- "account handling" operations -->
       <!-- Information request -->
       <element name="InfoRequest"/>
       <element name="InfoResponse">
        <complexType>
         <sequence>
          <element name="Status" type="string" minOccurs="0"/>
          <element name="ServerId" type="string"/>
          <element ref="sacred:AuthParams"/>
          <element ref="sacred:ProcessInfo" minOccurs="0"/>
         </sequence>
        </complexType>
       </element>
       <!-- Create Account Request -->
       <element name="CreateAccountRequest">
        <complexType>
         <sequence>
          <element name="UserId" type="string"/>
          <element ref="sacred:AuthInfo"/>
          <element ref="sacred:ProcessInfo" minOccurs="0"/>
         </sequence>
        </complexType>
       </element>
       <!-- remove account request -->
       <element name="RemoveAccountRequest">
        <complexType>
         <sequence>
          <element ref="sacred:ProcessInfo" minOccurs="0"/>
         </sequence>
        </complexType>
       </element>
       <!-- password change request -->
       <element name="ModifyAccountRequest">
        <complexType>
         <sequence>
          <element ref="sacred:AuthInfo"/>
          <element ref="sacred:ProcessInfo" minOccurs="0"/>
         </sequence>
        </complexType>
       </element>
       <!-- "run-time" operations -->



Farrell                     Standards Track                    [Page 18]

RFC 3767              Secure Credentials Protocol              June 2004


       <!-- DownLoad Request -->
       <element name="DownloadRequest">
        <complexType>
         <sequence>
          <element name="CredentialSelector" type="string"
            minOccurs="0"/>
          <element ref="sacred:ProcessInfo" minOccurs="0"/>
         </sequence>
        </complexType>
       </element>
       <!-- Download Response -->
       <element name="DownloadResponse">
        <complexType>
         <sequence>
          <element name="Status" type="string" minOccurs="0"/>
          <element name="Credential" type="sacred:CredentialType"
           maxOccurs="unbounded"/>
         </sequence>
        </complexType>
       </element>
       <!-- Upload request -->
       <element name="UploadRequest">
        <complexType>
         <sequence>
          <element name="Credential" type="sacred:CredentialType"/>
         </sequence>
        </complexType>
       </element>
       <element name="DeleteRequest">
         <complexType>
           <sequence>
             <choice>
               <sequence>
                 <element name="CredentialSelector" type="string"/>
                 <element name="LastModified" type="dateTime"
                       minOccurs="0"/>
               </sequence>
               <element name="All"/>
             </choice>
             <element ref="sacred:ProcessInfo" minOccurs="0"/>
           </sequence>
         </complexType>
       </element>
       <!-- Credential related structures -->
       <!-- A new ds:KeyInfo thing -->
       <element name="SacredPKCS15" type="base64Binary"/>
       <!-- credential -->
       <complexType name="CredentialType">



Farrell                     Standards Track                    [Page 19]

RFC 3767              Secure Credentials Protocol              June 2004


        <sequence>
         <element name="CredentialSelector" type="string"/>
         <element name="LastModified" type="dateTime"/>
         <element name="Payload" type="ds:KeyInfoType" minOccurs="0"/>
         <element name="TimeToLive" type="string" minOccurs="0"/>
         <element ref="sacred:ProcessInfo" minOccurs="0"/>
         <element ref="sacred:ClientInfo" minOccurs="0"/>
        </sequence>
       </complexType>

  </schema>

Appendix B: An Example of Tuning with BEEP

  Here is what tuning BEEP for authentication and confidentiality
  looks like using TLS and SASL's DIGEST-MD5:

  L: <wait for incoming connection>
  I: <open connection>

   ... each peer sends a greeting indicating the services that
      it offers ...

  L: RPY 0 0 . 0 233
  L: Content-Type: application/beep+xml
  L:
  L: <greeting>
  L:    <profile uri='http://iana.org/beep/SASL/DIGEST-MD5' />
  L:    <profile uri='http://iana.org/beep/TLS' />
  L:    <profile uri='http://iana.org/beep/sacred' />
  L: </greeting>
  L: END
  I: RPY 0 0 . 0 52
  I: Content-Type: application/beep+xml
  I:
  I: <greeting />
  I: END


   ... the initiator starts a channel for TLS and piggybacks a request
      to start the TLS negotiation ...

  I: MSG 0 1 . 52 149
  I: Content-Type: application/beep+xml
  I:
  I: <start number='1' serverName="sacred.example.org">
  I:    <profile uri='http://iana.org/beep/TLS'>
  I:        &lt;ready />



Farrell                     Standards Track                    [Page 20]

RFC 3767              Secure Credentials Protocol              June 2004


  I:    </profile>
  I: </start>
  I: END

   ... the listener creates the channel and piggybacks its readiness to
      start TLS ...

  L: RPY 0 1 . 233 112
  L: Content-Type: application/beep+xml
  L:
  L: <profile uri='http://iana.org/beep/TLS'>
  L:     &lt;proceed />
  L: </profile>
  L: END

   ... upon receiving the reply, the initiator starts up TLS ...

   ... successful transport security negotiation ...

   ... a new greeting is sent (cf., Section 9 of RFC 3080), note that
      the listener no longer advertises TLS (we're already running
      it)

  L: RPY 0 0 . 0 186
  L: Content-Type: application/beep+xml
  L:
  L: <greeting>
  L:    <profile uri='http://iana.org/beep/SASL/DIGEST-MD5' />
  L:    <profile uri='http://iana.org/beep/sacred' />
  L: </greeting>
  L: END
  I: RPY 0 0 . 0 52
  I: Content-Type: application/beep+xml
  I:
  I: <greeting />
  I: END


   ... the initiator starts a channel for DIGEST-MD5 and piggybacks
      initialization information for the mechanism ...

  I: MSG 0 1 . 52 178
  I: Content-Type: application/beep+xml
  I:
  I: <start number='1'>
  I:    <profile uri='http://iana.org/beep/SASL/DIGEST-MD5'>
  I:        &lt;blob> ... &lt;/blob>
  I:    </profile>



Farrell                     Standards Track                    [Page 21]

RFC 3767              Secure Credentials Protocol              June 2004


  I: </start>
  I: END

   ... the listener creates the channel and piggybacks a challenge ...

  L: RPY 0 1 . 186 137
  L: Content-Type: application/beep+xml
  L:
  L: <profile uri='http://iana.org/beep/SASL/DIGEST-MD5'>
  L:     &lt;blob> ... &lt;/blob>
  L: </profile>
  L: END

   ... the initiator sends a response to the challenge ...

  I: MSG 1 0 . 0 58
  I: Content-Type: application/beep+xml
  I:
  I: <blob> ... </blob>
  I: END

   ... the listener accepts the challenge and tells the initiator
      that it is now authenticated ...

  L: RPY 1 0 . 0 66
  L: Content-Type: application/beep+xml
  L:
  L: <blob status='complete' />
  L: END


   ... the initiator starts a channel for SACRED and piggybacks its
      initial SACRED request ...

  I: MSG 0 2 . 230 520
  I: Content-Type: application/beep+xml
  I:
  I: <start number='3'>
  I:    <profile uri='http://iana.org/beep/sacred' />
  I:        &lt;?xml version="1.0" encoding="UTF-8"?>
  I:        &lt;sacred:DownloadRequest
  I:          xmlns:sacred="urn:sacred-2002-12-19"
  I:          xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  I:          xsi:schemaLocation="urn:sacred-2002-12-19 sacred.xsd">
  I:          &lt;CredentialSelector>
  I:                      magnus-credentials&lt;/CredentialSelector>
  I:        &lt;/sacred:DownloadRequest>
  I: </start>



Farrell                     Standards Track                    [Page 22]

RFC 3767              Secure Credentials Protocol              June 2004


  I: END

   ... the listener creates the channel and piggybacks the response to
  the initial SACRED request

  L: RPY 0 2 . 323 805
  L: Content-Type: application/beep+xml
  L:
  L: <profile uri='http://iana.org/beep/sacred' />
  L:     &lt;?xml version="1.0" encoding="UTF-8"?>
  L:     &lt;sacred:DownloadResponse
  L:       xmlns:sacred="urn:sacred-2002-12-19"
  L:       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  L:       xsi:schemaLocation="urn:sacred-2002-12-19 sacred.xsd">
  L:       &lt;Status>Success&lt;/Status>
  L:       &lt;Credential>
  L:         &lt;CredentialSelector>
  L:              magnus-credential&lt;/CredentialSelector>
  L:         &lt;LastModified>2002-11-22T00:00:08Z&lt;/LastModified>
  L:         &lt;Payload>
  L:             &lt;sacred:SacredPKCS15
  L:               xmlns:sacred="urn:sacred-2002-12-19">GpM7
  L:             &lt;/sacred:SacredPKCS15>
  L:         &lt;/Payload>
  L:       &lt;/Credential>
  L:     &lt;/sacred:DownloadResponse>
  L: </profile>
  L: END

Appendix C: Provision SACRED using other Protocols

  SACRED may be implemented in a non-BEEP environment, provided that
  before any SACRED PDUs are sent, the application protocol must be
  protected according to the security mandates provided in Section 2.3.

  For example, if SACRED is provisioned as the payload of an
  application protocol that supports SASL and TLS, then the appropriate
  SASL and/or TLS negotiation must successfully occur before exchanging
  Sacred PDUs.

  Alternatively, if the application protocol doesn't support SASL, then
  one or more PDUs are defined to facilitate a SASL negotiation, and
  the appropriate negotiation must occur before exchanging Sacred PDUs.








Farrell                     Standards Track                    [Page 23]

RFC 3767              Secure Credentials Protocol              June 2004


Editor's Address

  Stephen Farrell,
  Distributed Systems Group,
  Computer Science Department,
  Trinity College Dublin,
  IRELAND
  Phone: +353-1-608-3070
  EMail: [email protected]










































Farrell                     Standards Track                    [Page 24]

RFC 3767              Secure Credentials Protocol              June 2004


Full Copyright Statement

  Copyright (C) The Internet Society (2004).  This document is subject
  to the rights, licenses and restrictions contained in BCP 78, and
  except as set forth therein, the authors retain all their rights.

  This document and the information contained herein are provided on an
  "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
  OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
  ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED,
  INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
  INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
  WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

  The IETF takes no position regarding the validity or scope of any
  Intellectual Property Rights or other rights that might be claimed to
  pertain to the implementation or use of the technology described in
  this document or the extent to which any license under such rights
  might or might not be available; nor does it represent that it has
  made any independent effort to identify any such rights.  Information
  on the procedures with respect to rights in RFC documents can be
  found in BCP 78 and BCP 79.

  Copies of IPR disclosures made to the IETF Secretariat and any
  assurances of licenses to be made available, or the result of an
  attempt made to obtain a general license or permission for the use of
  such proprietary rights by implementers or users of this
  specification can be obtained from the IETF on-line IPR repository at
  http://www.ietf.org/ipr.

  The IETF invites any interested party to bring to its attention any
  copyrights, patents or patent applications, or other proprietary
  rights that may cover technology that may be required to implement
  this standard.  Please address the information to the IETF at ietf-
  [email protected].

Acknowledgement

  Funding for the RFC Editor function is currently provided by the
  Internet Society.









Farrell                     Standards Track                    [Page 25]