Network Working Group                                        M. Crawford
Request for Comments: 2019                                      Fermilab
Category: Standards Track                                   October 1996



   A Method for the Transmission of IPv6 Packets over FDDI Networks

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.

Introduction

  This memo specifies the MTU and frame format for transmission of IPv6
  [IPV6] packets on FDDI networks, including a method for MTU
  determination in the presence of 802.1d bridges to other media.  It
  also specifies the method of forming IPv6 link-local addresses on
  FDDI networks and the content of the Source/Target Link-layer Address
  option used the the Router Solicitation, Router Advertisement,
  Neighbor Solicitation, and Neighbor Advertisement messages described
  in [DISC], when those messages are transmitted on an FDDI network.

Maximum Transmission Unit

  FDDI permits a frame length of 4500 octets (9000 symbols), including
  at least 22 octets (44 symbols) of Data Link encapsulation when
  long-format addresses are used.  Subtracting 8 octets of LLC/SNAP
  header, this would, in principle, allow the IPv6 packet in the
  Information field to be up to 4470 octets.  However, it is desirable
  to allow for the variable sizes and possible future extensions to the
  MAC header and frame status fields.  The default MTU size for IPv6
  packets on an FDDI network is therefore 4352 octets.  This size may
  be reduced by a Router Advertisement [DISC] containing an MTU option
  which specifies a smaller MTU, or by manual configuration of a
  smaller value on each node.  If a Router Advertisement is received
  with an MTU option specifying an MTU larger than the default or the
  manually configured value, that MTU option may be logged to system
  management but must be otherwise ignored.

  For purposes of this document, information received from DHCP is
  considered "manually configured".





Crawford                    Standards Track                     [Page 1]

RFC 2019         Transmission of IPv6 Packets Over FDDI     October 1996


Frame Format

  FDDI provides both synchronous and asynchronous transmission, with
  the latter class further subdivided by the use of restricted and
  unrestricted tokens.  Only asynchronous transmission with
  unrestricted tokens is required for FDDI interoperability.
  Accordingly, IPv6 packets shall be sent in asynchronous frames using
  unrestricted tokens.  The robustness principle dictates that nodes
  should be able to receive synchronous frames and asynchronous frames
  sent using restricted tokens.

  IPv6 packets are transmitted in LLC/SNAP frames, using long-format
  (48 bit) addresses.  The data field contains the IPv6 header and
  payload and is followed by the FDDI Frame Check Sequence, Ending
  Delimiter, and Frame Status symbols.

      +-------+                                               ^
      |  FC   |                                               |
      +-------+-------+-------+-------+-------+-------+       |
      |            Destination FDDI address           |       |
      +-------+-------+-------+-------+-------+-------+      FDDI
      |              Source FDDI address              |     header
      +-------+-------+-------+-------+-------+-------+       |
      | DSAP  | SSAP  |  CTL  |          OUI          |       |
      +-------+-------+-------+-------+-------+-------+       |
      |   Ethertype   |                                       v
      +-------+-------+-------+-------+-------+------+------+
      |            IPv6 header and payload ...              /
      +-------+-------+-------+-------+-------+------+------+

FDDI Header Fields:

FC          The Frame Code must be in the range 50 to 57 hexadecimal,
           inclusive, with the three low order bits indicating the
           frame priority.  The Frame Code should be in the range 51 to
           57 hexadecimal, inclusive, for reasons given in the next
           section.

DSAP, SSAP  Both the DSAP and SSAP fields shall contain the value AA
           hexadecimal, indictating SNAP encapsulation.

CTL         The Control field shall be set to 03 hexadecimal, indicating
           Unnumbered Information.

OUI         The Organizationally Unique Identifier shall be set to
           000000 hexadecimal.





Crawford                    Standards Track                     [Page 2]

RFC 2019         Transmission of IPv6 Packets Over FDDI     October 1996


Ethertype   The ethernet protocol type ("ethertype") shall be set to the
           value 86DD hexadecimal.

Interaction with Bridges

  802.1d MAC bridges which connect different media, for example
  Ethernet and FDDI, have become very widespread.  Some of them do IPv4
  packet fragmentation and/or support IPv4 Path MTU discovery [PMTU],
  many others do not, or do so incorrectly.  Use of IPv6 in a bridged
  mixed-media environment should not depend on support from MAC
  bridges.

  For correct operation when mixed media are bridged together, the
  smallest MTU of all the media must be advertised by routers in an MTU
  option.  If there are no routers present, this MTU must be manually
  configured in each node which is connected to a medium with larger
  default MTU.  Multicast packets on such a bridged network must not be
  larger than the smallest MTU of any of the bridged media.  Often, the
  subnetwork topology will support larger unicast packets to be
  exchanged between certain pairs of nodes.  To take advantage of
  high-MTU paths when possible, nodes transmitting IPv6 on FDDI should
  implement the following simple mechanism for "FDDI adjacency
  detection".

  A node which implements FDDI adjacency detection and has it enabled
  on an FDDI interface must set a non-zero LLC priority in all Neighbor
  Advertisement, Neighbor Solicitation and, if applicable, Router
  Advertisement frames transmitted on that interface.  (In IEEE 802
  language, the user_priority parameter of the M_UNITDATA.request
  primitive must not be zero.) If FDDI adjacency detection has been
  disabled on an FDDI interface, the priority field of those frames
  must be zero.

  Note that an IPv6 frame which originated on an Ethernet, or traversed
  an Ethernet, before being translated by an 802.1d bridge and
  delivered to a node's FDDI interface will have zero in the priority
  field, as required by [BRIDGE].  (There's a fine point here: a
  conforming bridge may provide a management-settable Outbound User
  Priority parameter for each port.  However, the author is unaware of
  any product that provides this optional capability and, in any case,
  the default value for the parameter is zero.)










Crawford                    Standards Track                     [Page 3]

RFC 2019         Transmission of IPv6 Packets Over FDDI     October 1996


  If a node N1 receives, in an FDDI frame with a non-zero LLC priority,
  a valid Router Advertisement, Neighbor Advertisement, or Neighbor
  Solicitation from a node N2, then N1 may send unicast IPv6 packets to
  N2 with sizes up to the default IPv6 FDDI MTU (4352 octets),
  regardless of any smaller MTU configured manually or received in a
  Router Advertisement MTU option.  N2 may be the IPv6 destination or
  the next hop router to the destination.

  Nodes implementing FDDI adjacency detection must provide a
  configuration option to disable the mechanism.  This option may be
  used when a smaller MTU is desired for reasons other than mixed-media
  bridging.  By default, FDDI adjacency detection should be enabled.

  The only contemplated use of the LLC priority field of the FC octet
  is to aid in per-destination MTU determination.  It would be
  sufficient for that purpose to require only that Router
  Advertisements, Neighbor Advertisements, and Neighbor Solicitations
  sent on FDDI always have non-zero priority.  However, it may be
  simpler or more useful to transmit all IPv6 packets on FDDI with
  non-zero priority.

Stateless Autoconfiguration and Link-Local Addresses

  The address token [CONF] for an FDDI interface is the interface's
  built-in 48-bit IEEE 802 address, in canonical bit order and with the
  octet in the same order in which they would appear in the header of
  an ethernet frame.  (The individual/group bit is in the first octet
  and the OUI is in the first three octets.) A different MAC address
  set manually or by software should not be used as the address token.

  An IPv6 address prefix used for stateless autoconfiguration of an
  FDDI interface must be 80 bits in length.

  The IPv6 Link-local address [AARCH] for an FDDI interface is formed
  by appending the interface's IEEE 802 address to the 80-bit prefix
  FE80::.

     +-------+-------+-------+-------+-------+-------+------+------+
     |  FE      80      00      00      00      00      00     00  |
     +-------+-------+-------+-------+-------+-------+------+------+
     |  00      00   |                  FDDI Address               |
     +-------+-------+-------+-------+-------+-------+------+------+









Crawford                    Standards Track                     [Page 4]

RFC 2019         Transmission of IPv6 Packets Over FDDI     October 1996


Address Mapping -- Unicast

  The procedure for mapping IPv6 addresses into FDDI link-layer
  addresses is described in [DISC].  The Source/Target Link-layer
  Address option has the following form when the link layer is FDDI.

     +-------+-------+-------+-------+-------+-------+------+------+
     | Type  |Length |                 FDDI Address                |
     +-------+-------+-------+-------+-------+-------+------+------+

Option fields:

Type        1 for Source Link-layer address.
           2 for Target Link-layer address.

Length      1 (in units of 8 octets).

FDDI Address
           The 48 bit FDDI IEEE 802 address, in canonical bit order.
           This is the address the interface currently responds to, and
           may be different from the built-in address used as the
           address token.

Address Mapping -- Multicast

  An IPv6 packet with a multicast destination address DST is
  transmitted to the FDDI multicast address whose first two octets are
  the value 3333 hexadecimal and whose last four octets are the last
  four octets of DST, ordered from more to least significant.

            +-------+-------+-------+-------+-------+-------+
            |  33   |  33   | DST13 | DST14 | DST15 | DST16 |
            +-------+-------+-------+-------+-------+-------+


















Crawford                    Standards Track                     [Page 5]

RFC 2019         Transmission of IPv6 Packets Over FDDI     October 1996


Security Considerations

  Security considerations are not addressed in this memo.

Acknowledgments

  Erik Nordmark contributed to the method for interaction with bridges.

References

  [AARCH] Hinden, and S. Deering, "IP Version 6 Addressing
          Architecture", RFC 1884, December 1995.

  [BRIDGE]ISO/IEC 10038 : 1993 [ANSI/IEEE Std 802.1D] Media access
          control (MAC) bridges.

  [CONF] Thomson, S., and T. Narten, "IPv6 Stateless Address
         Autoconfiguration", RFC 1971, August 1996.

  [DISC] Narten, T., Nordmark, E., and W. Simpson, "Neighbor Discovery
         for IP Version 6 (IPv6), RFC 1970, August 1996.

  [IPV6] Deering, S., and R. Hinden, "Internet Protocol, Version 6
         (IPv6) Specification", RFC 1883, August 1996.

  [PMTU] Mogul, J., and S. Deering, "Path MTU Discovery", RFC 1191,
         November 1990.

Author's Address

  Matt Crawford
  Fermilab MS 368
  PO Box 500
  Batavia, IL 60510
  USA

  Phone: +1 708 840-3461

  EMail: [email protected]












Crawford                    Standards Track                     [Page 6]