P2P over LAN: Point-to-Point Operation over LAN in Link State Routing Protocols

Network Working Group                                       N. Shen, Ed.
Request for Comments: 5309                                 Cisco Systems
Category: Informational                                    A. Zinin, Ed.
                                                          Alcatel-Lucent
                                                            October 2008


                   Point-to-Point Operation over LAN
                    in Link State Routing Protocols

Status of This Memo

   This memo provides information for the Internet community.  It does
   not specify an Internet standard of any kind.  Distribution of this
   memo is unlimited.

Abstract

   The two predominant circuit types used by link state routing
   protocols are point-to-point and broadcast.  It is important to
   identify the correct circuit type when forming adjacencies, flooding
   link state database packets, and representing the circuit
   topologically.  This document describes a simple mechanism to treat
   the broadcast network as a point-to-point connection from the
   standpoint of IP routing.

1.  Introduction

   Point-to-point and broadcast are the two predominant circuit types
   used by link state routing protocols such as IS-IS [ISO10589]
   [RFC1195] and OSPF [RFC2328] [RFC5340].  They are treated differently
   with respect to establishing neighbor adjacencies, flooding link
   state information, representing the topology, and calculating the
   Shortest Path First (SPF) and protocol packets.  The most important
   differences are that broadcast circuits utilize the concept of a
   designated router and are represented topologically as virtual nodes
   in the network topology graph.

   Compared with broadcast circuits, point-to-point circuits afford more
   straightforward IGP operation.  There is no designated router
   involved, and there is no representation of the pseudonode or network
   Link State Advertisement (LSA) in the link state database.  For IS-
   IS, there also is no periodic database synchronization.  Conversely,
   if there are more than two routers on the LAN media, the traditional
   view of the broadcast circuit will reduce the routing information in
   the network.

   When there are only two routers on the LAN, it makes more sense to
   treat the connection between the two routers as a point-to-point
   circuit.  This document describes the mechanism to allow link state
   routing protocols to operate using point-to-point connections over a
   LAN under this condition.  Some implications related to forwarding IP
   packets on this type of circuit are also discussed.  We will refer to
   this as a p2p-over-lan circuit in this document.

1.1.  Terminology

   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 RFC 2119 [RFC2119].

2.  Motivation

   Even though a broadcast circuit is meant to handle more than two
   devices, there are cases where only two routers are connected over
   either the physical or logical LAN segment:

      1. The media itself is being used for point-to-point operation
         between two routers.  This is mainly for long-haul operation.
      2. There are only two routers on the physical LAN.
      3. There are only two routers on a virtual LAN (vLAN).

   In any of the above cases, the link state routing protocols will
   normally still treat the media as a broadcast circuit.  Hence, they
   will have the overhead involved with protocol LAN operation without
   the benefits of reducing routing information and optimized flooding.

   Being able to treat a LAN as a point-to-point circuit provides the
   benefit of reduction in the amount of information routing protocols
   must carry and manage.  DR/DIS (Designated Router / Designated
   Intermediate System) election can be omitted.  Flooding can be done
   as in p2p links without the need for using "LSA reflection" by the DR
   in OSPF or periodic Complete Sequence Number Packets (CSNPs) in IS-
   IS.

   Also, if a broadcast segment wired as a point-to-point link can be
   treated as a point-to-point link, only the connection between the two
   routers would need to be advertised as a topological entity.

   Even when there are multiple routers on the LAN, an ISP may want to
   sub-group the routers into multiple vLANs, since this allows them to
   assign different costs to IGP neighbors.  When there are only two
   routers in some of the vLANs, this LAN can be viewed by the IGP as a
   mesh of point-to-point connections.

   The IP unnumbered configuration is widely used in networks.  It
   enables IP processing on a point-to-point interface without an
   explicit IP address.  The IP unnumbered interface can "borrow" the IP
   address of another interface on the node.  The advantages of
   unnumbered point-to-point links are obvious in the current IP
   addressing environment where addresses are a scarce resource.  The
   unnumbered interface can also be applied over p2p-over-lan circuits.
   Separating the concept of network type from media type will allow
   LANs, e.g., ethernet, to be unnumbered and realize the IP address
   space savings.  Another advantage is in simpler network management
   and configuration.  In the case of an IPv6 network, a link local
   address used in IS-IS [RFC5308] and OSPFv3 [RFC5340] serves the same
   purpose.

3.  IP Multi-Access Subnets

   When an IP network includes multi-access segments, each segment is
   usually assigned a separate subnet, and each router connected to it
   is assigned a distinct IP address within that subnet.  The role of
   the IP address assigned to a multi-access interface can be outlined
   as follows:

      1. Source IP address - The interface address can be used by the
         router as the source IP address in locally originated IP
         packets that are destined for that subnet or have a best path
         next hop on that subnet.

      2. Destination IP address - The interface address can be used by
         other devices in the network as a destination address for
         packets to router applications (examples include telnet, SMTP,
         TFTP, OSPF, BGP, etc).

      3. Next-hop identifier - If other routers connected to the same
         segment need to forward traffic through the router, the
         corresponding routes in their routing tables will include the
         router's interface IP address.  This address will be used to
         find the router's MAC (Media Access Control) address using the
         ARP/ND (Address Resolution Protocol / Neighbor Discovery)
         protocol.  Effectively, the interface IP addresses help other
         routers find the data-link layer details that are required to
         specify the destination of the encapsulating data-link frame
         when it is sent on the segment.

   The IP addressing scheme includes an option that allows the
   administrators to not assign any subnets to point-to-point links
   (links connecting only two devices and using protocols like PPP,
   SLIP, or HDLC for IP encapsulation).  This is possible because the
   routers do not need next-hop identifiers on point-to-point links
   (there is only one destination for any transmission), and an
   interface-independent IP address can be used as the source and
   destination.  Using the unnumbered option for a point-to-point link
   essentially makes it a purely topological entity used only to reach
   other destinations.

4.  Point-to-Point Connection over LAN Media

   The idea is very simple: provide a configuration mechanism to inform
   the IGP that the circuit is type point-to-point, irrespective of the
   physical media type.  For the IGP, this implies that it will send
   protocol packets with the appropriate point-to-point information, and
   it expects to receive protocol packets as they would be received on a
   point-to-point circuit.  Over LAN media, the MAC header must contain
   the correct multicast MAC address to be received by the other side of
   the connection.  For vLAN environments, the MAC header must also
   contain the proper vLAN ID.

   In order to allow LAN links used to connect only two routers to be
   treated as unnumbered point-to-point interfaces, the MAC address
   resolution and nexthop IP address issues need to be addressed.

4.1.  Operation of IS-IS

   This p2p-over-lan circuit extension for IS-IS is only concerned with
   pure IP routing and forwarding operation.

   Since physically the circuit is a broadcast one, the IS-IS protocol
   packets need to have MAC addresses for this p2p-over-lan circuit.
   From a link-layer point of view, those packets are IS-IS LAN packets.
   The Multi-destination address including AllISs, AllL1ISs, and
   AllL2ISs, defined in [ISO10589], can be used for link-layer
   encapsulation; the use of AllISs is recommended.

   The circuit needs to have IP address(es), and the p2p IS-IS Hello
   (IIH) over this circuit MUST include the IP interface address(es) as
   defined in [RFC1195].  The IPv4 address(es) included in the IIHs is
   either the IP address assigned to the interface in the case of a
   numbered interface or the interface-independent IP address in the
   case of an unnumbered interface.  The IPv6 addresses are link-local
   IPv6 address(es) [RFC5308].

4.2.  Operation of OSPF and OSPFv3

   OSPF and OSPFv3 [RFC5340] routers supporting the capabilities
   described herein should support an additional interface configuration
   parameter specifying the interface topology type.  For a LAN (i.e.,
   broadcast-capable) interface, the interface may be viewed as a
   point-to-point interface.  Both routers on the LAN will simply join
   the AllSPFRouters multicast group and send all OSPF packets with a
   destination address of AllSPFRouters.  AllSPFRouters is 224.0.0.5 for
   OSPF and FF02::5 for OSPFv3.  This is identical to operation over a
   physical point-to-point link as described in Sections 8.1 and 8.2 of
   [RFC2328].

4.3.  ARP and ND

   Unlike a normal point-to-point IGP circuit, the IP nexthop for the
   routes using this p2p-over-lan circuit as an outbound interface is
   not optional.  The IP nexthop address has to be a valid interface or
   internal address on the adjacent router.  This address is used by a
   local router to obtain the MAC address for IP packet forwarding.  The
   ARP process has to be able to resolve the internal IPv4 address used
   for the unnumbered p2p-over-lan circuits.  For the ARP implementation
   (which checks that the subnet of the source address of the ARP
   request matches the local interface address), this check needs to be
   relaxed for the unnumbered p2p-over-lan circuits.  The
   misconfiguration detection is handled by the IGPs and is described in
   Section 4.5.  In the IPv6 case, the ND resolves the MAC for the
   link-local address on the p2p-over-lan circuit, which is part of the
   IPv6 neighbor discovery process [RFC4861].

4.4.  Other MAC Address Resolution Mechanisms

   In more general cases, while p2p-over-lan circuit is used as an
   unnumbered link, other MAC address resolution mechanisms are needed
   for IP packet forwarding; for example, if link state IGP is not
   configured over this p2p-over-lan link, or if the mechanism described
   in Section 4.3 is not possible.  The following techniques can be used
   to acquire the MAC address and/or the next-hop IP address of the
   remote device on an unnumbered point-to-point LAN link.

      1. Static configuration.  A router can be statically configured
         with the MAC address that should be used as the destination MAC
         address when sending data out of the interface.

      2. MAC address gleaning.  If a dynamic routing protocol is running
         between the routers connected to the link, the MAC address of
         the remote device can be taken from a data-link frame carrying
         a packet of the corresponding routing protocol.

4.5.  Detection of Misconfiguration

   With this p2p-over-lan extension, the difference between a LAN and a
   point-to-point circuit can be made purely by configuration.  It is
   important to implement the mechanisms for early detection of
   misconfiguration.

   If the circuit is configured as the point-to-point type and receives
   LAN hello packets, the router MUST discard the incoming packets; if
   the circuit is a LAN type and receives point-to-point hello packets,
   it MUST discard the incoming packets.  If the system ID or the router
   ID of an incoming hello packet does not match the system ID or the
   router ID for an established adjacency over a p2p-over-lan circuit,
   the packet MUST be discarded.  Furthermore, if OSPF hello suppression
   (as described in [RFC1793]) is active for the adjacency, the hello
   suppression MUST be terminated for a period of RouterIntervalSeconds.
   After this interval, either the neighbor adjacency will time out and
   an adjacency may be formed with a neighbor with a different router
   ID, or hello suppression may be renegotiated.  The implementation
   should offer logging and debugging information of the above events.

5.  Compatibility Considerations

   Both routers on a LAN must support the p2p-over-lan extension and
   both must have the LAN segment configured as a p2p-over-lan circuit
   for successful operation.  Both routers SHOULD support at least one
   of the above listed methods for mapping IP addresses on the link to
   MAC address.  If a proprietary method of IP address to MAC address
   resolution is used by one router, both routers must be capable of
   using the same method.  Otherwise, the link should be configured as a
   standard LAN link, with traditional IGP LAN models used.

6.  Scalability and Deployment Considerations

   While there is advantage to using this extension on the LANs that are
   connected back to back or only contain two routers, there are trade
   offs when modeling a LAN as multiple vLANs and using this extension
   since one does sacrifice the inherent scalability benefits of multi-
   access networks.  In general, it will increase the link state
   database size, the amount of packets flooded, and the route
   calculation overhead.

   Deployment of the described technique brings noticeable benefits from
   the perspective of IP address usage: the network management and the
   router configuration.  Note, however, that use of the IP unnumbered
   option for point-to-point LAN links inherits the same problems as
   those present for serial links, i.e., not being able to ping or
   monitor a specific interface between routers.

7.  Security Considerations

   This document does not introduce any new security issues to IS-IS,
   OSPF, ARP, or ND.  Implementations may have 'source address subnet
   checks' that need to be relaxed as described in Section 4.3.  These
   are used to manage misconfigurations, not so much to secure ARP -- if
   an attacker would be attached to the LAN, (s)he could pick a subnet-
   wise correct address as well.

   If one router on a link thinks that a LAN should be either broadcast
   or p2p-over-lan, and the other router has a different opinion, the
   adjacencies will never form, as specified in Section 4.5.  There are
   no fallbacks at either end to resolve the situation, except by a
   manual configuration change.

8.  Acknowledgments

   The authors would like to acknowledge the following individuals (in
   alphabetical order by last name): Pedro Marques, Christian Martin,
   Danny McPherson, Ajay Patel, Jeff Parker, Tony Przygienda, Alvaro
   Retana, and Pekka Savola.

9.  Normative References

   [ISO10589] ISO, "Intermediate System to Intermediate System intra-
              domain routeing information exchange protocol for use in
              conjunction with the protocol for providing the
              connectionless-mode network service (ISO 8473)",
              International Standard 10589:2002, Second Edition, 2002.

   [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
              dual environments", RFC 1195, December 1990.

   [RFC1793]  Moy, J., "Extending OSPF to Support Demand Circuits", RFC
              1793, April 1995.

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

   [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.

   [RFC4861]  Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
              "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
              September 2007.

   [RFC5308]  Hopps, C., "Routing IPv6 with IS-IS", RFC 5308, October
              2008.

   [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
              for IPv6", RFC 5340, July 2008.

Contributors

   The following individuals are the authors that contributed to the
   contents of this document.

   Acee Lindem
   Cisco Systems
   7025 Kit Creek Road
   Research Triangle Park, NC  27709
   USA
   EMail: acee@cisco.com

   Jenny Yuan
   Cisco Systems
   225 West Tasman Drive
   San Jose, CA 95134
   USA
   EMail: jenny@cisco.com

   Russ White
   Cisco Systems, Inc.
   7025 Kit Creek Rd.
   Research Triangle Park, NC 27709
   EMail: riw@cisco.com

   Stefano Previdi
   Cisco Systems, Inc.
   De Kleetlaan 6A
   1831 Diegem - Belgium
   EMail: sprevidi@cisco.com

Editors' Addresses

   Naiming Shen
   Cisco Systems
   225 West Tasman Drive
   San Jose, CA  95134
   USA
   EMail: naiming@cisco.com

   Alex Zinin
   Alcatel-Lucent
   750D Chai Chee Rd, #06-06
   Technopark@ChaiChee
   Singapore 469004

   EMail: alex.zinin@alcatel-lucent.com

Full Copyright Statement

   Copyright (C) The IETF Trust (2008).

   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, THE IETF TRUST 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-ipr@ietf.org.