Network Working Group                                          J. Allen
Request for Comments: 2653                         WebTV Networks, Inc.
Category: Standards Track                                      P. Leach
                                                             Microsoft
                                                            R. Hedberg
                                                             Catalogix
                                                           August 1999


                       CIP Transport 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.

Abstract

  This document specifies three protocols for transporting CIP
  requests, responses and index objects, utilizing TCP, mail, and HTTP.
  The objects themselves are defined in [CIP-MIME] and the overall CIP
  architecture is defined in [CIP-ARCH].

1.   Protocol

  In this section, the actual protocol for transmitting CIP index
  objects and maintaining the mesh is presented. While companion
  documents ([CIP-ARCH] and [CIP-MIME]) describe the concepts involved
  and the formats of the CIP MIME objects, this document is the
  authoritative definition of the message formats and transfer
  mechanisms of CIP used over TCP, HTTP and mail.

1.1  Philosophy

  The philosophy of the CIP protocol design is one of building-block
  design. Instead of relying on bulky protocol definition tools, or
  ad-hoc text encodings, CIP draws on existing, well understood
  Internet technologies like MIME, RFC-822, Whois++, FTP, and SMTP.
  Hopefully this will serve to ease implementation and consensus





Allen, et al.               Standards Track                     [Page 1]

RFC 2653                CIP Transport Protocols              August 1999


  building. It should also stand as an example of a simple way to
  leverage existing Internet technologies to easily implement new
  application-level services.

1.2  Conventions

  The key words "MUST" and "MAY" in this document are to be interpreted
  as described in "Key words for use in RFCs to Indicate Requirement
  Levels" [KEYWORDS].

  Formal syntax is defined using ABNF [ABNF].

  In examples octets sent by the sender-CIP are preceded by ">>> " and
  those sent by the receiver-CIP by "<<< ".

2  MIME message exchange mechanisms

  CIP relies on interchange of standard MIME messages for all requests
  and replies. These messages are passed over a bidirectional, reliable
  transport system. This document defines transport over reliable
  network streams (via TCP), via HTTP, and via the Internet mail
  infrastructure.

  The CIP server which initiates the connection (conventionally
  referred to as a client) will be referred to below as the sender-CIP.
  The CIP server which accepts a sender-CIP's incoming connection and
  responds to the sender-CIP's requests is called a receiver-CIP.

2.1  The Stream Transport

  CIP messages are transmitted over bi-directional TCP connections via
  a simple text protocol. The transaction can take place over any TCP
  port, as specified by the mesh configuration. There is no "well known
  port" for CIP transactions. All configuration information in the
  system must include both a hostname and a port.

  All sender-CIP actions (including requests, connection initiation,
  and connection finalization) are acknowledged by the receiver-CIP
  with a response code. See section 2.1.1 for the format of these
  codes, a list of the responses a CIP server may generate, and the
  expected sender-CIP action for each.

  In order to maintain backwards compatibility with existing Whois++
  servers, CIPv3 sender-CIPs MUST first verify that the newer protocol
  is supported. They do this by sending the following illegal Whois++
  system command: "# CIP-Version: 3<cr><lf>". On existing Whois++
  servers implementing version 1 and 2 of CIP, this results in a 500-
  series response code, and the server terminates the connection.  If



Allen, et al.               Standards Track                     [Page 2]

RFC 2653                CIP Transport Protocols              August 1999


  the server implements CIPv3, it MUST instead respond with response
  code 300. Future versions of CIP can be correctly negotiated using
  this technique with a different string (i.e. "CIP-Version: 4"). An
  example of this short interchange is given below.

  Note: If a sender-CIP can safely assume that the server implements
  CIPv3, it may choose to send the "# CIP-Version: 3" string and
  immediately follow it with the CIPv3 request. This optimization,
  useful only in known homogeneous CIPv3 meshes, avoids waiting for the
  roundtrip inherent in the negotiation.

  Once a sender-CIP has successfully verified that the server supports
  CIPv3 requests, it can send the request, formatted as a MIME message
  with Mime-Version and Content-Type headers (only), using the network
  standard line ending: "<cr><lf>".

  Cip-Req        = Req-Hdrs CRLF Req-Body

  Req-Hdrs       = *( Version-Hdr | Req-Cntnt-Hdr )
  Req-Body       = Body     ; format of request body as in [CIP-MIME]

  Body           = Data CRLF "." CRLF
  Data           =          ; data with CRLF "." CRLF replaced by
                            ; CRLF ".." CRLF

  Version-Hdr    = "Mime-Version:" "1.0" CRLF
  Req-Cntnt-Hdr  = "Content-Type:" Req-Content CRLF
  Req-Content    =          ; format is specified in [CIP-MIME]

  Cip-Rsp        = Rsp-Code CRLF [ Rsp-Hdrs CRLF Rsp-Body ]
                    [ Indx-Cntnt-Hdr CRLF Index-Body ]
  Rsp-Code       = DIGIT DIGIT DIGIT Comment
  Comment        =          ; any chars except CR and LF
  Rsp-Hdrs       = *( Version-Hdr | Rsp-Cntnt-Hdr )
  Rsp-Cntnt-Hdr  = "Content-Type:" Rsp-Content CRLF
  Rsp-Content    =          ; format is specified in [CIP-MIME]
  Rsp-Body       = Body     ; format of response body as in [CIP-MIME]

  Indx-Cntnt-Hdr = "Content-Type:" Indx-Obj-Type CRLF
  Indx-Obj-Type  =          ; any registered index object's MIME-type
                            ; the format is specified in [RFC2045]
  Index-Body     = Body     ; format defined in each index
                            ; specifications

  CRLF           =  CR LF   ; Internet standard newline
  CR             =  %x0D    ; carriage return
  LF             =  %x0A    ; linefeed
  DIGIT          =  %x30-39



Allen, et al.               Standards Track                     [Page 3]

RFC 2653                CIP Transport Protocols              August 1999


  The message is terminated using SMTP-style message termination. The
  data is sent octet-for-octet, except when the pattern
  <cr><lf>1*["."]<cr><lf> is seen, in which case one more period is
  added.

  When the data is finished, the octet pattern "<cr><lf>.<cr><lf>" is
  transmitted to the receiver-CIP.

  On the receiver-CIP's side, the reverse transformation is applied,
  and the message read consists of all bytes up to, but not including,
  the terminating pattern.

  In response to the request, the receiver-CIP sends a response code,
  from either the 200, 400, or 500 series. The receiver-CIP then
  processes the request and replies, if necessary, with a MIME message.
  This reply is also delimited by an SMTP-style message terminator.

  After responding with a response code, the receiver-CIP MUST prepare
  to read another request message, resetting state to the point when
  the sender-CIP has just verified the CIP version. If the sender-CIP
  is finished making requests, it may close the connection. In response
  the receiver-CIP MUST abort reading the message and prepare for a new
  sender-CIP connection (resetting its state completely).

  An example is given below. It is again worth reiterating that the
  command format is defined in [CIP-MIME] whereas the message body is
  defined in each index object definition. In this example the index
  object definition in [CIP-TIO] will be used. Line endings are
  explicitly shown in anglebrackets; newlines in this text are added
  only for readability. Comments occur in curly-brackets.

     { sender-CIP connects to receiver-CIP }
  <<< % 220 Example CIP server ready<cr><lf>
  >>> # CIP-Version: 3<cr><lf>
  <<< % 300 CIPv3 OK!<cr><lf>
  >>> Mime-Version: 1.0<cr><lf>
  >>> Content-type: application/index.cmd.datachanged; type=
  >>> x-tagged-index-1; dsi=1.2.752.17.5.10<cr><lf>
  >>> <cr><lf>
  >>> updatetype: incremental tagbased<cr><lf>
  >>> thisupdate: 855938804<cr><lf>
  >>> lastupdate: 855940000<cr><lf>
  >>> .<cr><lf>
  <<< % 200 Good MIME message received
  >>> MIME-Version: 1.0<cr><lf>
  >>> Content-Type: application/index.obj.tagged;
  >>> dsi=1.2.752.17.5.10;
  >>> base-uri="ldap://ldap.umu.se/dc=umu,dc=se"<cr><lf>



Allen, et al.               Standards Track                     [Page 4]

RFC 2653                CIP Transport Protocols              August 1999


  >>> <cr><lf>
  >>> version: x-tagged-index-1<cr><lf>
  >>> updatetype: incremental<cr><lf>
  >>> lastupdate: 855940000<cr><lf>
  >>> thisupdate: 855938804<cr><lf>
  >>> BEGIN IO-schema<cr><lf>
  >>> cn: TOKEN<cr><lf>
  >>> sn: FULL<cr><lf>
  >>> title: FULL<cr><lf>
  >>> END IO-Schema<cr><lf>
  >>> BEGIN Update Block<cr><lf>
  >>> BEGIN Old<cr><lf>
  >>> title: 3/testpilot<cr><lf>
  >>> END Old<cr><lf>
  >>> BEGIN New<cr><lf>
  >>> title: 3/chiefpilot<cr><lf>
  >>> END New<cr><lf>
  >>> END Update Block<cr><lf>
  >>> .<cr><lf>
  <<< % 200 Good MIME message received
     { Sender-CIP shuts down socket for writing }
  <<< % 222 Connection closing in response to sender-CIP shutdown
     { receiver-CIP closes its side, resets, and awaits a
       new sender-CIP }

  An example of an unsuccessful version negotiation looks like this:

     { sender-CIP connects to receiver-CIP }
  <<< % 220 Whois++ server ready<cr><lf>
  >>> # CIP-Version: 3<cr><lf>
  <<< % 500 Syntax error<cr><lf>
     { server closes connection }

  The sender-CIP may attempt to retry using version 1 or 2 protocol.
  Sender-CIP may cache results of this unsuccessful negotiation to
  avoid later attempts.

2.1.1     Transport specific response codes

  The following response codes are used with the stream transport:

  Code  Suggested description     Sender-CIP action
        text

  200   MIME request received     Expect no output, continue session
        and processed             (or close)





Allen, et al.               Standards Track                     [Page 5]

RFC 2653                CIP Transport Protocols              August 1999


  201   MIME request received     Read a response, delimited by SMTP-
        and processed, output     style message delimiter.
        follows

  220   Initial server banner     Continue with Whois++ interaction,
        message                   or attempt CIP version negotiation.

  222   Connection closing (in    Done with transaction.
        response to sender-CIP
        close)

  300   Requested CIP version     Continue with CIP transaction, in
        accepted                  the specified version.

  400   Temporarily unable to     Retry at a later time. May be used
        process request           to indicate that the server does not
                                  currently have the resources
                                  available to accept an index.

  500   Bad MIME message format   Retry with correctly formatted MIME

  501   Unknown or missing        Retry with correct CIP command
        request in
        application/index.cmd

  502   Request is missing        Retry with correct CIP attributes.
        required CIP attributes

  520   Aborting connection for   Alert local administrator.
        some unexpected reason

  530   Request requires valid    Sign the request, if possible, and
        signature                 retry. Otherwise, report problem to
                                  the administrator.

  531   Request has invalid       Report problem to the administrator.
        signature

  532   Cannot check signature    Alert local administrator, who should
                                  cooperate with remote administrator
                                  tp diagnose and resolve the problem.
                                  (Probably missing a public key.)









Allen, et al.               Standards Track                     [Page 6]

RFC 2653                CIP Transport Protocols              August 1999


2.2  Internet mail infrastructure as transport

  As an alternative to TCP streams, CIP transactions can take place
  over the existing Internet mail infrastructure. There are two
  motivations for this feature of CIP. First, it lowers the barriers to
  entry for leaf servers. When the need for a full TCP implementation
  is relaxed, leaf nodes (which, by definition, only send index
  objects) can consist of as little as a database and an indexing
  program (possibly written in a very high level language) to
  participate in the mesh.

  Second, it keeps with the philosophy of making use of existing
  Internet technology. The MIME messages used for requests and
  responses are, by definition of the MIME specification, suitable for
  transport via the Internet mail infrastructure. With a few simple
  rules, we open up an entirely different way to interact with CIP
  servers which choose to implement this transport. See Protocol
  Conformance, below, for details on what options server implementers
  have about supporting the various transports.

  The basic rhythm of request/response is maintained when using the
  mail transport. The following sections clarify some special cases
  which need to be considered for mail transport of CIP objects. In
  general, all mail protocols and mail format specifications
  (especially MIME Security Multiparts) can be used with the CIP mail
  transport.

2.2.1     CIP-Version negotiation

  Since no information on which CIP-version is in use is present in the
  MIME message, this information has to be carried in the mailheader.
  Therefore CIP requests sent using the mail transport MUST include a
  CIP-version headerline, to be registered according to [MHREG].
  The format of this line is:

  DIGIT       =  %x30-39
  number      =  1*DIGIT
  cipversion  =  "CIP-Version:" <sp> number["." number]

2.2.2     Return path

  When CIP transactions take place over a bidirectional stream, the
  return path for errors and results is implicit. Using mail as a
  transport introduces difficulties to the recipient, because it's not
  always clear from the headers exactly where the reply should go,
  though in practice there are some heuristics used by MUA's.





Allen, et al.               Standards Track                     [Page 7]

RFC 2653                CIP Transport Protocols              August 1999


  CIP solves this problem by fiat. CIP requests sent using the mail
  transport MUST include a Reply-To header as specified by RFC-822.
  Any mail received for processing by a CIP server implementing the
  mail transport without a Reply-To header MUST be ignored, and a
  message should be logged for the local administrator. The receiver
  MUST not attempt to reply with an error to any address derived from
  the incoming mail.

  If there are no circumstances under which a response is to be sent to
  a CIP request, the sender should include a Reply-To header with the
  address "<>" in it.  Receivers MUST never attempt to  send replies to
  that address, as it is defined to be invalid (both here, and by the
  BNF grammar in RFC-822). It should be noted that, in general, it is a
  bad idea to turn off error reporting in this way. However, in the
  simplest case of an index pushing program, this MAY be a desirable
  simplification.

2.3  HTTP transport

  HTTP MAY also be used to transport CIP objects, since they are just
  MIME objects. A transaction is performed by using the POST method to
  send an application/index.cmd and returning an
  application/index.response or an application/index.obj in the HTTP
  reply. The URL that is the target of the post is a configuration
  parameter of the CIP-sender to CIP-receiver relationship.

  Example:

     { the client opens the connection and sends a POST }
  >>> POST / HTTP/1.1<cr><lf>
  >>> Host: cip.some.corp<cr><lf>
  >>> Content-type: application/index.cmd.noop<cr><lf>
  >>> Date: Thu, 6 Jun 1997 18:16:03 GMT<cr><lf>
  >>> Content-Length: 2<cr><lf>
  >>> Connection: close<cr><lf>
  >>> <cr><lf>
     { the server processes the request }
  <<< HTTP/1.1 204 No Content<cr><lf>
     { the server closes the connection }

  In addition to leveraging the security capabilities that come with
  HTTP, there are other HTTP features that MAY be useful in a CIP
  context. A CIP client MAY use the Accept-Charset and Accept-Language
  HTTP headers to express a desire to retrieve an index in a particular
  character set or natural language. It MAY use the Accept-Encoding
  header to (e.g.) indicate that it can handle compressed responses,
  which the CIP server MAY send in conjunction with the Transfer-
  Encoding header. It MAY use the If-Modified-Since header to prevent



Allen, et al.               Standards Track                     [Page 8]

RFC 2653                CIP Transport Protocols              August 1999


  wasted transmission of an index that has not changed since the last
  poll. A CIP server can use the Retry-After header to request that the
  client retry later when the server is less busy.

3.   Security Considerations

  There are two levels at which the index information can be protected;
  the first is by use of the technology available for securing MIME
  [MIME-SEC] objects, and secondly by using the technology available
  for securing the transport.

  When it comes to transport the stream transport can be protected by
  the use of TLS [TLS] . For HTTP the Security is handled by using HTTP
  Basic Authentication [RFC 2616], HTTP Message Digest Authentication
  [RFC2617] or SSL/TLS. Extra protection for the SMTP exchange can be
  achieve by the use of Secure SMTP over TLS [SMTPTLS].

4.   References

  [RFC 2045] Freed, N. and N. Borenstein, "Multipurpose Internet Mail
             Extensions (MIME) Part One: Format of Internet Message
             Bodies", RFC 2045, November 1996.

  [RFC 2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H.,
             Masinter, L., Leach, P. and T. Berners-Lee, "Hypertext
             Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999.

  [RFC 2617] Franks, J., Hallam-Baker, P., Hostetler, J., Lawrence, S.,
             Leach, P., Luotonen, A. and L. Stewart, "HTTP
             Authentication: Basic and Digest Access Authentication",
             RFC 2617, June 1999.

  [CIP-ARCH] Allen, J. and M. Mealling, "The Architecture of the Common
             Indexing Protocol (CIP)", RFC 2651, August 1999.

  [CIP-MIME] Allen, J. and M. Mealling, "MIME Object Definitions for
             the Common Indexing Protocol (CIP)", RFC 2652, August
             1999.

  [ABNF]     Crocker, D. and P. Overell, "Augmented BNF for Syntax
             Specifications: ABNF", RFC 2234, November 1997.

  [CIP-TIO]  Hedberg, R., Greenblatt, B., Moats, R. and M. Wahl, "A
             Tagged Index Object for use in the Common Indexing
             Protocol", RFC 2654, August 1999.

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



Allen, et al.               Standards Track                     [Page 9]

RFC 2653                CIP Transport Protocols              August 1999


  [MIME-SEC] Galvin, J., Murphy, S., Crocker, S. and N. Freed,
             "Security Multiparts for MIME: Multipart/Signed and
             Multipart/Encrypted", RFC 1847, October 1995.

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

  [SMTPTLS]  Hoffman, P., "SMTP Service Extension for Secure SMTP over
             TLS", RFC 2487, January 1999.

  [MHREG]    Jacob, P., "Mail and Netnews Header Registration
             Procedure", Work in Progress.

5.   Authors' Addresses

  Jeff R. Allen
  246 Hawthorne St.
  Palo Alto, CA  94301

  EMail: [email protected]


  Paul J. Leach
  Microsoft
  1 Microsoft Way
  Redmond, WA 98052

  EMail: [email protected]


  Roland Hedberg
  Catalogix
  Dalsveien 53
  0775 Oslo
  Norway

  EMail: [email protected]














Allen, et al.               Standards Track                    [Page 10]

RFC 2653                CIP Transport Protocols              August 1999


6.  Full Copyright Statement

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

  This document and translations of it may be copied and furnished to
  others, and derivative works that comment on or otherwise explain it
  or assist in its implementation may be prepared, copied, published
  and distributed, in whole or in part, without restriction of any
  kind, provided that the above copyright notice and this paragraph are
  included on all such copies and derivative works.  However, this
  document itself may not be modified in any way, such as by removing
  the copyright notice or references to the Internet Society or other
  Internet organizations, except as needed for the purpose of
  developing Internet standards in which case the procedures for
  copyrights defined in the Internet Standards process must be
  followed, or as required to translate it into languages other than
  English.

  The limited permissions granted above are perpetual and will not be
  revoked by the Internet Society or its successors or assigns.

  This document and the information contained herein is provided on an
  "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
  TASK FORCE DISCLAIMS 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.

Acknowledgement

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



















Allen, et al.               Standards Track                    [Page 11]