RFC 7040






Internet Engineering Task Force (IETF)                            Y. Cui
Request for Comments: 7040                                         J. Wu
Category: Informational                                            P. Wu
ISSN: 2070-1721                                      Tsinghua University
                                                              O. Vautrin
                                                        Juniper Networks
                                                                  Y. Lee
                                                                 Comcast
                                                           November 2013


                  Public IPv4-over-IPv6 Access Network

Abstract



   This document describes a mechanism called Public 4over6, which is
   designed to provide IPv4 Internet connectivity over an IPv6 access
   network using global IPv4 addresses.  Public 4over6 was developed in
   the IETF and is in use in some existing deployments but is not
   recommended for new deployments.  Future deployments of similar
   scenarios should use Lightweight 4over6.  Public 4over6 follows the
   Hub and Spoke softwire model and uses an IPv4-in-IPv6 tunnel to
   forward IPv4 packets over an IPv6 access network.  The
   bidirectionality of the IPv4 communication is achieved by explicitly
   allocating global non-shared IPv4 addresses to end users and by
   maintaining IPv4-IPv6 address binding on the border relay.  Public
   4over6 aims to provide uninterrupted IPv4 services to users, like
   Internet Content Providers (ICPs), etc., while an operator makes the
   access network transition to an IPv6-only access network.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Not all documents
   approved by the IESG are a candidate for any level of Internet
   Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc7040.






Cui, et al.                   Informational                     [Page 1]

RFC 7040                      Public 4over6                November 2013


Copyright Notice



   Copyright (c) 2013 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1. Introduction ....................................................2
   2. Terminology .....................................................4
   3. Scenario and Use Cases ..........................................4
   4. Public 4over6 Address Provisioning ..............................6
      4.1. Basic Provisioning Steps ...................................6
      4.2. Public IPv4 Address Allocation .............................7
   5. 4over6 CE Behavior ..............................................7
   6. 4over6 BR Behavior ..............................................8
   7. Fragmentation and Reassembly ....................................9
   8. DNS .............................................................9
   9. Security Considerations ........................................10
   10. Contributors ..................................................11
   11. References ....................................................12
      11.1. Normative References .....................................12
      11.2. Informative References ...................................12

1.  Introduction



   When operators make the access network transition to an IPv6-only
   access network, they must continue to provide IPv4 services to their
   users to access IPv4 contents.  IPv4 connectivity is required when
   communicating with the IPv4-only Internet.  This document describes a
   mechanism called Public 4over6 for providing IPv4 connectivity over a
   native IPv6-only access network.  This memo focuses on interactions
   between Public 4over6 elements as well as the deployment
   architecture.








Cui, et al.                   Informational                     [Page 2]

RFC 7040                      Public 4over6                November 2013


   Public 4over6 is in active deployment in some environments,
   particularly in China Next Generation Internet (CNGI) and China
   Education and Research Network 2 (CERNET2), but it is not recommended
   for new deployments.  Documenting this approach is intended to
   benefit users and operators of existing deployments as well as
   readers of other IPv4-over-IPv6 documents.

   In addition to Public 4over6 and its deployment architecture as
   described in this memo, the IETF is currently working on a more
   generic solution called Lightweight 4over6 [SOFTWIRE-LW46], which is
   classified as a binding approach in the Unified IPv4-in-IPv6 Softwire
   Customer Premises Equipment (CPE) [SOFTWIRE-CPE].  Lightweight 4over6
   covers both sharing and non-sharing global IPv4 addresses in the Hub
   and Spoke model.  Future deployments should use [SOFTWIRE-LW46].

   Public 4over6 utilizes the IPv4-in-IPv6 tunnel technique defined in
   [RFC2473], which enables IPv4 datagrams to traverse through native
   IPv6 networks.  IPv4 nodes connect to the Tunnel Entry-Point Node and
   Tunnel Exit-Point Node to communicate over the IPv6-only network.
   Therefore, the Internet Service Providers (ISPs) can run an IPv6-only
   infrastructure instead of a fully dual-stack network as well as avoid
   the need to deploy scarce IPv4 address resources throughout the
   network.

   This mechanism focuses on providing full end-to-end transparency to
   the user side.  Therefore, global IPv4 addresses are expected to be
   provisioned to end users, and carrier-side address translation can be
   avoided.  Furthermore, global non-shared IPv4 addresses are
   preferable to shared IPv4 addresses, so that user-side address
   translation is not necessary either.  It is important, in particular,
   to users that are required to run their applications on an IP
   protocol different from TCP and UDP (e.g., IPsec, L2TP) or on certain
   well-known TCP/UDP ports (e.g., HTTP, SMTP).  For many ISPs that are
   actually capable of provisioning non-shared unique IPv4 addresses,
   the mechanism provides a pure, suitable solution.

   Another focus of this mechanism is deployment and operational
   flexibility.  Public 4over6 allows IPv4 and IPv6 address
   architectures to be totally independent of each other; the end user's
   IPv4 address is not embedded in its IPv6 address.  Therefore, IPv4
   address planning has no implication for IPv6 address planning.
   Operators can manage the IPv4 address resources in a flat,
   centralized manner.  This requires that the tunnel concentrator
   [RFC4925] maintain the binding between an IPv4 address and an IPv6
   address, i.e., maintaining per-subscriber binding state.






Cui, et al.                   Informational                     [Page 3]

RFC 7040                      Public 4over6                November 2013


   The mechanism follows the Hub and Spoke softwire model [RFC4925] and
   uses IPv4-in-IPv6 tunneling as the basic data-plane method.  Global
   non-shared IPv4 addresses are allocated from the ISP to end hosts or
   CPEs over an IPv6 network.  Simultaneously, the binding between the
   allocated IPv4 address and the end user's IPv6 address is maintained
   on the tunnel concentrator for encapsulation usage.

2.  Terminology



   Public 4over6:  A per-subscriber, stateful IPv4-in-IPv6 tunnel
      mechanism.  Public 4over6 supports bidirectional communication
      between the global IPv4 Internet and IPv4 hosts or customer
      networks via an IPv6 access network by leveraging IPv4-in-IPv6
      tunneling [RFC2473] and global IPv4 address allocation over IPv6.
      The term 'Public' means the allocated IPv4 address is globally
      routable.

   Full IPv4 address:  An IPv4 address that is not shared by multiple
      users.  The user with this IPv4 address has full access to all the
      available TCP/UDP ports, including the well-known TCP/UDP ports.

   4over6 Customer Edge (CE):  A device functioning as the Customer Edge
      equipment in a Public 4over6 environment.  A 4over6 CE can be
      either a dual-stack capable host or a dual-stack CPE device, both
      of which have a tunnel interface to support IPv4-in-IPv6
      encapsulation.  In the former case, the host supports both IPv4
      and IPv6 stacks but its uplink is IPv6 only.  In the latter case,
      the CPE has an IPv6 interface connecting to the ISP network and an
      IPv4 or dual-stack interface connecting to the customer network;
      hosts in the customer network can be IPv4 only or dual stack.

   4over6 Border Relay (BR):  A router deployed in the edge of the
      operator's IPv6 access network that supports IPv4-in-IPv6 tunnel
      termination.  A 4over6 BR is a dual-stack router that connects to
      both the IPv6 access network and the IPv4 Internet.  The 4over6 BR
      can also work as a DHCPv4-over-IPv6 [DHCPv4-IPv6] server/relay for
      assigning and distributing global IPv4 addresses to 4over6 CEs.

3.  Scenario and Use Cases



   The general Public 4over6 scenario is shown in Figure 1.  Users in an
   IPv6 network take IPv6 as their native service.  Some users are end
   hosts that face the ISP network directly, while others are in private
   networks behind CPEs, such as a home Local Area Network (LAN), an
   enterprise network, etc.  The ISP network is IPv6 only rather than
   dual stack, which means the ISP cannot provide native IPv4 service to
   users.  In order to support legacy IPv4 transport, some routers on




Cui, et al.                   Informational                     [Page 4]

RFC 7040                      Public 4over6                November 2013


   the carrier side are dual stack and are connected to the IPv4
   Internet.  These routers act as 4over6 BRs.  Network users that
   require IPv4 connectivity obtain it through these routers.

                        +-------------------------+
                        |    IPv6 ISP Network     |
                        |                         |
                     +------+                     |
                     |4over6|Host             +-------+   +-----------+
                     |  CE  |=================|       |   |           |
                     +------+                 |       |   |           |
                        |                     |4over6 |   |   IPv4    |
   +--------------+  +------+  IPv4-in-IPv6   |  BR   |---| Internet  |
   |   Customer   |  |4over6|                 |       |   |           |
   | Private IPv4 |--|  CE  |=================|       |   |           |
   |   Network    |  |      |CPE              +-------+   +-----------+
   +--------------+  +------+                     |
                        |                         |
                        |                         |
                        +-------------------------+

                     Figure 1: Public 4over6 Scenario

   Public 4over6 can be applicable in several use cases.  If an ISP that
   switches to IPv6 still has plenty of global IPv4 address resources,
   it can deploy Public 4over6 to provide transparent IPv4 service for
   all its customers.  If the ISP does not have enough IPv4 addresses,
   it can deploy Dual-Stack Lite [RFC6333] as the basic IPv4-over-IPv6
   service.  Along with Dual-Stack Lite, Public 4over6 can be deployed
   as a value-added service, overcoming the service degradation caused
   by the Carrier Grade NAT (CGN).  An IPv4 application server is a
   typical high-end user of Public 4over6.  Using a full, global IPv4
   address brings significant advantages in this case and is important
   for Internet Content Providers (ICPs) making the transition to IPv6:

   o  The DNS registration can be direct, using a dedicated address;

   o  Accessing the application service can be straightforward, with no
      translation involved;

   o  There will be no need to provide NAT traversal mechanisms for
      incoming traffic, and no special handling is required for the
      well-known TCP/UDP ports.








Cui, et al.                   Informational                     [Page 5]

RFC 7040                      Public 4over6                November 2013


4.  Public 4over6 Address Provisioning



4.1.  Basic Provisioning Steps



   Figure 2 shows the basic provisioning steps for Public 4over6.

         4over6                  DHCPv6          4over6         DHCPv4
           CE                    Server            BR           Server
           |Assign IPv6 Addr/Pref +|               |              |
           |  BR's IPv6 Addr Info  |               |              |
           |<----------------------|               |              |
           |     DHCPv6/Other      |               |              |
          WAN                                      |              |
       IPv6 Configure                              |              |
           |                                       |              |
           | Assign Public IPv4 Addr (DHCPv4 over v6/Static Conf) |
           |<--------------------------------------|<-------------|
           |                                       | IPv4-IPv6    |
           |                                       | Binding SYN  |
          Tunnel                                   |
       IPv4 Configure                        Binding Update
           |                                       |
           |          IPv4-in-IPv6 Tunnel          |
           |<------------------------------------->|
           |                                       |

               Figure 2: Public 4over6 Address Provisioning

   The main steps are:

   o  IPv6 address/prefix is provisioned to 4over6 CE, along with
      knowledge of 4over6 BR's IPv6 address, using DHCPv6 or other
      means.

   o  4over6 CE configures its WAN interface with a globally unique IPv6
      address, which is a result of IPv6 provisioning, including DHCPv6,
      Stateless Address Autoconfiguration (SLAAC), or manual
      configuration.

   o  IPv4 address is provisioned to 4over6 CE by DHCPv4 over IPv6 or
      static configuration.

   o  4over6 BR obtains the IPv4 and IPv6 addresses of the 4over6 CE
      using information provided by the DHCPv4 server.







Cui, et al.                   Informational                     [Page 6]

RFC 7040                      Public 4over6                November 2013


   o  4over6 CE configures its tunnel interface as a result of IPv4
      provisioning.

   o  4over6 BR updates the IPv4-IPv6 address-binding table according to
      the address-binding information acquired from the DHCPv4 server.

4.2.  Public IPv4 Address Allocation



   Usually, each CE is provisioned with one global IPv4 address.
   However, it is possible that a CE would require an IPv4 prefix.  The
   key problem here is the mechanism for IPv4 address provisioning over
   IPv6 networks.

   There are two possibilities: DHCPv4 over IPv6, and static
   configuration.  Public 4over6 supports both these methods.  DHCPv4
   over IPv6 allows DHCPv4 messages to be transported in IPv6 rather
   than IPv4; therefore, the DHCPv4 process can be performed over an
   IPv6 network between the BR and the relevant CE. [DHCPv4-IPv6]
   describes the DHCP protocol extensions needed to support this
   operation.  For static configuration, Public 4over6 users and ISP
   operators negotiate beforehand to authorize the IPv4 address(es).
   Then the tunnel interface and the address binding are configured by
   the user and the ISP, respectively.

   While regular users would probably opt for DHCPv4 over IPv6, the
   static configuration is particularly applicable in two cases: for
   application servers, which require a stable IPv4 address; and for
   enterprise networks, which usually require an IPv4 prefix rather than
   one single address.  (Note that DHCPv4 does not support prefix
   allocation.)

5.  4over6 CE Behavior



   A CE is provisioned with IPv6 before the Public 4over6 process.  It
   also learns the BR's IPv6 address beforehand.  This IPv6 address can
   be configured using a variety of methods, ranging from an out-of-band
   mechanism, manual configuration, or via a DHCPv6 option.  In order to
   guarantee interoperability, the CE element implements the AFTR-Name
   DHCPv6 option defined in [RFC6334].

   A CE supports DHCPv4 over IPv6 [DHCPv4-IPv6] to dynamically acquire
   an IPv4 address over IPv6 and assign it to the IPv4-in-IPv6 tunnel
   interface.  The CE regards the BR as its DHCPv4-over-IPv6
   server/relay, which is used to obtain its global IPv4 address
   allocation; its IPv6 address is learned by the CE as described above.






Cui, et al.                   Informational                     [Page 7]

RFC 7040                      Public 4over6                November 2013


   A CE also supports static configuration of the tunnel interface.  In
   the case of prefix provisioning, the tunnel interface is assigned
   with the well-known IPv4 address defined in Section 5.7 of [RFC6333],
   rather than using an address from the prefix.  If the CE has multiple
   IPv6 addresses on its WAN interface, it uses one of the IPv6
   addresses for DHCPv4 over IPv6 or negotiation of static
   configuration.  The CE then uses the same IPv6 address for data-plane
   encapsulation.

   A CE performs IPv4-in-IPv6 encapsulation and decapsulation on the
   tunnel interface.  When sending out an IPv4 packet, it performs the
   encapsulation using the IPv6 address of the 4over6 BR as the IPv6
   destination address and its own IPv6 address as the IPv6 source
   address.  The decapsulation on the 4over6 CE is simple.  When
   receiving an IPv4-in-IPv6 packet, the CE just removes the IPv6 header
   and either hands it to a local upper layer or forwards it to the
   customer network according to the IPv4 destination address.

   A CE runs a regular IPv4 Network Address and Port Translation (NAPT)
   for its customer network when it is provisioned with one single IPv4
   address.  In that case, the assigned IPv4 address of the tunnel
   interface would be the external IPv4 address of the NAPT.  Then the
   CE performs IPv4 private-to-public translation before encapsulation
   of IPv4 packets from the customer network and IPv4 public-to-private
   translation after decapsulation of IPv4-in-IPv6 packets.

   IPv4 NAPT is not necessary when the CE is provisioned with an IPv4
   prefix.  In this case, detailed customer network planning is out of
   scope for this document.

   The 4over6 CE supports backward compatibility with DS-Lite.  A CE can
   employ the well-known IPv4 address for the Basic Bridging BroadBand
   (B4) element [RFC6333] and switch to Dual-Stack Lite for IPv4
   communications if it can't get a global IPv4 address from the DHCPv4
   server (for instance, if the DHCPv4-over-IPv6 process fails or the
   DHCPv4 server refuses to allocate a global IPv4 address to it, etc.).

6.  4over6 BR Behavior



   The 4over6 BR maintains the bindings between the CE IPv6 address and
   CE IPv4 address (prefixes).  The bindings are used to provide the
   correct encapsulation destination address for inbound IPv4 packets
   and also to validate the IPv6-IPv4 source of the outbound IPv4-in-
   IPv6 packets.







Cui, et al.                   Informational                     [Page 8]

RFC 7040                      Public 4over6                November 2013


   The BR acquires the binding information through the IPv4 address
   provisioning process.  For static configuration, the operator
   manually configures the BR using the binding information obtained
   through negotiation with the customer.  As for DHCPv4 over IPv6,
   there are multiple possibilities, which are deployment-specific:

   o  The BR can be co-located with the DHCPv4-over-IPv6 server.  Then
      the synchronization happens within the BR.  It installs a binding
      when sending out an ACK for a DHCP lease and deletes it when the
      lease expires or a DHCP RELEASE message is received.

   o  The BR can play the role of IPv6-Transport Relay Agent (TRA) as
      described in [DHCPv4-IPv6] and snoop for the DHCPv4 ACK and
      RELEASE messages as well as keep a timer for each binding
      according to the DHCP lease time.

   On the IPv6 side, the BR decapsulates IPv4-in-IPv6 packets coming
   from 4over6 CEs.  It removes the IPv6 header of every IPv4-in-IPv6
   packet and forwards it to the IPv4 Internet.  Before the
   decapsulation, the BR checks the inner IPv4 source address against
   the outer IPv6 source address by matching such a binding entry in the
   binding table.  If no binding is found, the BR silently drops the
   packet.  On the IPv4 side, the BR encapsulates the IPv4 packets
   destined to 4over6 CEs.  When performing the IPv4-in-IPv6
   encapsulation, the BR uses its own IPv6 address as the IPv6 source
   address and uses the IPv4 destination address in the packet to look
   up the IPv6 destination address in the address-binding table.  After
   the encapsulation, the BR sends the IPv6 packet on its IPv6 interface
   to reach a CE.

   The BR supports the hairpinning of traffic between two CEs by
   performing decapsulation and re-encapsulation of packets.

   In cases where the BR manages the global IPv4 address pool, the BR
   advertises the routing information of IPv4 addresses to the IPv4
   Internet.

7.  Fragmentation and Reassembly



   The same considerations as those described in Sections 5.3 and 6.3 of
   [RFC6333] are taken into account for the CE and the BR, respectively.

8.  DNS



   The procedure described in Sections 5.5 and 6.4 of [RFC6333] is
   followed by the CE and the BR, respectively.





Cui, et al.                   Informational                     [Page 9]

RFC 7040                      Public 4over6                November 2013


9.  Security Considerations



   The 4over6 BR implements methods to limit service only to registered
   customers.  On the control plane, the BR allocates IPv4 addresses
   only to registered customers.  The BR can filter on the IPv6 source
   addresses of incoming DHCP requests and only respond to the ones that
   are conveyed by registered IPv6 source addresses.  But this doesn't
   work in situations where multi-homing is present.  In the networks
   where Public 4over6 is deployed, multi-homing is disallowed to avoid
   this issue.

   Alternatively, the BR can filter out the unregistered CE's requests
   during the DHCP process.  For data packets, the BR does ingress
   filtering by looking up addresses in the IPv4-IPv6 address-binding
   table for the related matches as described in Section 6.

   In the case of fallback to DS-Lite, security considerations in
   Section 11 of [RFC6333] are followed.

































Cui, et al.                   Informational                    [Page 10]

RFC 7040                      Public 4over6                November 2013


10.  Contributors



   The following are those who have made contributions to the effort:

      Huiling Zhao
      China Telecom
      Room 502, No.118, Xizhimennei Street
      Beijing  100035
      P.R.China
      Phone: +86-10-58552002
      Email: zhaohl@ctbri.com.cn


      Chongfeng Xie
      China Telecom
      Room 708, No.118, Xizhimennei Street
      Beijing  100035
      P.R.China
      Phone: +86-10-58552116
      Email: xiechf@ctbri.com.cn


      Qiong Sun
      China Telecom
      Room 708, No.118, Xizhimennei Street
      Beijing  100035
      P.R.China
      Phone: +86-10-58552936
      Email: sunqiong@ctbri.com.cn


      Qi Sun
      Tsinghua University
      Beijing  100084
      P.R.China
      Phone: +86-10-62785822
      Email: sunqi@csnet1.cs.tsinghua.edu.cn


      Chris Metz
      Cisco Systems
      3700 Cisco Way
      San Jose, CA  95134
      USA
      Email: chmetz@cisco.com






Cui, et al.                   Informational                    [Page 11]

RFC 7040                      Public 4over6                November 2013


11.  References



11.1.  Normative References



   [RFC2473]  Conta, A. and S. Deering, "Generic Packet Tunneling in
              IPv6 Specification", RFC 2473, December 1998.

   [RFC4925]  Li, X., Dawkins, S., Ward, D., and A. Durand, "Softwire
              Problem Statement", RFC 4925, July 2007.

   [RFC6333]  Durand, A., Droms, R., Woodyatt, J., and Y. Lee, "Dual-
              Stack Lite Broadband Deployments Following IPv4
              Exhaustion", RFC 6333, August 2011.

   [RFC6334]  Hankins, D. and T. Mrugalski, "Dynamic Host Configuration
              Protocol for IPv6 (DHCPv6) Option for Dual-Stack Lite",
              RFC 6334, August 2011.

11.2.  Informative References



   [DHCPv4-IPv6]
              Cui, Y., Wu, P., Wu, J., Lemon, T., and Q. Sun, "DHCPv4
              over IPv6 Transport", Work in Progress, October 2013.

   [SOFTWIRE-CPE]
              Boucadair, M., Farrer, I., Perreault, S., Ed., and S.
              Sivakumar, Ed., "Unified IPv4-in-IPv6 Softwire CPE", Work
              in Progress, May 2013.

   [SOFTWIRE-LW46]
              Cui, Y., Sun, Q., Boucadair, M., Tsou, T., Lee, Y., and I.
              Farrer, "Lightweight 4over6: An Extension to the DS-Lite
              Architecture", Work in Progress, November 2013.


















Cui, et al.                   Informational                    [Page 12]

RFC 7040                      Public 4over6                November 2013


Authors' Addresses



   Yong Cui
   Tsinghua University
   Beijing  100084
   P.R.China
   Phone: +86-10-6260-3059
   EMail: yong@csnet1.cs.tsinghua.edu.cn


   Jianping Wu
   Tsinghua University
   Beijing  100084
   P.R.China
   Phone: +86-10-6278-5983
   EMail: jianping@cernet.edu.cn


   Peng Wu
   Tsinghua University
   Beijing  100084
   P.R.China
   Phone: +86-10-6278-5822
   EMail: pengwu.thu@gmail.com


   Olivier Vautrin
   Juniper Networks
   1194 N. Mathilda Avenue
   Sunnyvale, CA  94089
   USA
   EMail: Olivier@juniper.net


   Yiu L. Lee
   Comcast
   One Comcast Center
   Philadelphia, PA  19103
   USA
   EMail: yiu_lee@cable.comcast.com











Cui, et al.                   Informational                    [Page 13]