RFC 4456






Network Working Group                                           T. Bates
Request for Comments: 4456                                       E. Chen
Obsoletes: 2796, 1966                                      Cisco Systems
Category: Standards Track                                     R. Chandra
                                                           Sonoa Systems
                                                              April 2006


                         BGP Route Reflection:
            An Alternative to Full Mesh Internal BGP (IBGP)


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 (2006).

Abstract



   The Border Gateway Protocol (BGP) is an inter-autonomous system
   routing protocol designed for TCP/IP internets.  Typically, all BGP
   speakers within a single AS must be fully meshed so that any external
   routing information must be re-distributed to all other routers
   within that Autonomous System (AS).  This represents a serious
   scaling problem that has been well documented with several
   alternatives proposed.

   This document describes the use and design of a method known as
   "route reflection" to alleviate the need for "full mesh" Internal BGP
   (IBGP).

   This document obsoletes RFC 2796 and RFC 1966.












Bates, et al.               Standards Track                     [Page 1]

RFC 4456                  BGP Route Reflection                April 2006


Table of Contents



   1. Introduction ....................................................2
   2. Specification of Requirements ...................................2
   3. Design Criteria .................................................3
   4. Route Reflection ................................................3
   5. Terminology and Concepts ........................................4
   6. Operation .......................................................5
   7. Redundant RRs ...................................................6
   8. Avoiding Routing Information Loops ..............................6
   9. Impact on Route Selection .......................................7
   10. Implementation Considerations ..................................7
   11. Configuration and Deployment Considerations ....................7
   12. Security Considerations ........................................8
   13. Acknowledgements ...............................................9
   14. References .....................................................9
      14.1. Normative References ......................................9
      14.2. Informative References ....................................9
   Appendix A: Comparison with RFC 2796 ..............................10
   Appendix B: Comparison with RFC 1966 ..............................10

1.  Introduction



   Typically, all BGP speakers within a single AS must be fully meshed
   and any external routing information must be re-distributed to all
   other routers within that AS.  For n BGP speakers within an AS that
   requires to maintain n*(n-1)/2 unique Internal BGP (IBGP) sessions.
   This "full mesh" requirement clearly does not scale when there are a
   large number of IBGP speakers each exchanging a large volume of
   routing information, as is common in many of today's networks.

   This scaling problem has been well documented, and a number of
   proposals have been made to alleviate this [2,3].  This document
   represents another alternative in alleviating the need for a "full
   mesh" and is known as "route reflection".  This approach allows a BGP
   speaker (known as a "route reflector") to advertise IBGP learned
   routes to certain IBGP peers.  It represents a change in the commonly
   understood concept of IBGP, and the addition of two new optional
   non-transitive BGP attributes to prevent loops in routing updates.

   This document obsoletes RFC 2796 [6] and RFC 1966 [4].

2.  Specification of Requirements



   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 [7].




Bates, et al.               Standards Track                     [Page 2]

RFC 4456                  BGP Route Reflection                April 2006


3.  Design Criteria



   Route reflection was designed to satisfy the following criteria.

      o  Simplicity

         Any alternative must be simple to configure and easy to
         understand.

      o  Easy Transition

         It must be possible to transition from a full-mesh
         configuration without the need to change either topology or AS.
         This is an unfortunate management overhead of the technique
         proposed in [3].

      o  Compatibility

         It must be possible for noncompliant IBGP peers to continue to
         be part of the original AS or domain without any loss of BGP
         routing information.

   These criteria were motivated by operational experiences of a very
   large and topology-rich network with many external connections.

4.  Route Reflection



   The basic idea of route reflection is very simple.  Let us consider
   the simple example depicted in Figure 1 below.

                   +-------+        +-------+
                   |       |  IBGP  |       |
                   | RTR-A |--------| RTR-B |
                   |       |        |       |
                   +-------+        +-------+
                         \            /
                     IBGP \   ASX    / IBGP
                           \        /
                            +-------+
                            |       |
                            | RTR-C |
                            |       |
                            +-------+

                    Figure 1: Full-Mesh IBGP

   In ASX, there are three IBGP speakers (routers RTR-A, RTR-B, and
   RTR-C).  With the existing BGP model, if RTR-A receives an external



Bates, et al.               Standards Track                     [Page 3]

RFC 4456                  BGP Route Reflection                April 2006


   route and it is selected as the best path it must advertise the
   external route to both RTR-B and RTR-C.  RTR-B and RTR-C (as IBGP
   speakers) will not re-advertise these IBGP learned routes to other
   IBGP speakers.

   If this rule is relaxed and RTR-C is allowed to advertise IBGP
   learned routes to IBGP peers, then it could re-advertise (or reflect)
   the IBGP routes learned from RTR-A to RTR-B and vice versa.  This
   would eliminate the need for the IBGP session between RTR-A and RTR-B
   as shown in Figure 2 below.

                  +-------+        +-------+
                  |       |        |       |
                  | RTR-A |        | RTR-B |
                  |       |        |       |
                  +-------+        +-------+
                        \            /
                    IBGP \   ASX    / IBGP
                          \        /
                           +-------+
                           |       |
                           | RTR-C |
                           |       |
                           +-------+

                Figure 2: Route Reflection IBGP

   The route reflection scheme is based upon this basic principle.

5.  Terminology and Concepts



   We use the term "route reflection" to describe the operation of a BGP
   speaker advertising an IBGP learned route to another IBGP peer.  Such
   a BGP speaker is said to be a "route reflector" (RR), and such a
   route is said to be a reflected route.

   The internal peers of an RR are divided into two groups:

      1) Client peers

      2) Non-Client peers

   An RR reflects routes between these groups, and may reflect routes
   among client peers.  An RR along with its client peers form a
   cluster.  The Non-Client peer must be fully meshed but the Client
   peers need not be fully meshed.  Figure 3 depicts a simple example
   outlining the basic RR components using the terminology noted above.




Bates, et al.               Standards Track                     [Page 4]

RFC 4456                  BGP Route Reflection                April 2006


                 / - - - - - - - - - - - - -  -
                 |           Cluster           |
                   +-------+        +-------+
                 | |       |        |       |  |
                   | RTR-A |        | RTR-B |
                 | |Client |        |Client |  |
                   +-------+        +-------+
                 |       \           /         |
                    IBGP  \         / IBGP
                 |         \       /           |
                           +-------+
                 |         |       |           |
                           | RTR-C |
                 |         |  RR   |           |
                           +-------+
                 |           /   \             |
                  - - - - - /- - -\- - - - - - /
                     IBGP  /       \ IBGP
                  +-------+         +-------+
                  | RTR-D |  IBGP   | RTR-E |
                  |  Non- |---------|  Non- |
                  |Client |         |Client |
                  +-------+         +-------+

                     Figure 3: RR Components

6.  Operation



   When an RR receives a route from an IBGP peer, it selects the best
   path based on its path selection rule.  After the best path is
   selected, it must do the following depending on the type of peer it
   is receiving the best path from

      1) A route from a Non-Client IBGP peer:

         Reflect to all the Clients.

      2) A route from a Client peer:

         Reflect to all the Non-Client peers and also to the Client
         peers.  (Hence the Client peers are not required to be fully
         meshed.)

   An Autonomous System could have many RRs.  An RR treats other RRs
   just like any other internal BGP speakers.  An RR could be configured
   to have other RRs in a Client group or Non-client group.





Bates, et al.               Standards Track                     [Page 5]

RFC 4456                  BGP Route Reflection                April 2006


   In a simple configuration, the backbone could be divided into many
   clusters.  Each RR would be configured with other RRs as Non-Client
   peers (thus all the RRs will be fully meshed).  The Clients will be
   configured to maintain IBGP session only with the RR in their
   cluster.  Due to route reflection, all the IBGP speakers will receive
   reflected routing information.

   It is possible in an Autonomous System to have BGP speakers that do
   not understand the concept of route reflectors (let us call them
   conventional BGP speakers).  The route reflector scheme allows such
   conventional BGP speakers to coexist.  Conventional BGP speakers
   could be members of either a Non-Client group or a Client group.
   This allows for an easy and gradual migration from the current IBGP
   model to the route reflection model.  One could start creating
   clusters by configuring a single router as the designated RR and
   configuring other RRs and their clients as normal IBGP peers.
   Additional clusters can be created gradually.

7.  Redundant RRs



   Usually, a cluster of clients will have a single RR.  In that case,
   the cluster will be identified by the BGP Identifier of the RR.
   However, this represents a single point of failure so to make it
   possible to have multiple RRs in the same cluster, all RRs in the
   same cluster can be configured with a 4-byte CLUSTER_ID so that an RR
   can discard routes from other RRs in the same cluster.

8.  Avoiding Routing Information Loops



   When a route is reflected, it is possible through misconfiguration to
   form route re-distribution loops.  The route reflection method
   defines the following attributes to detect and avoid routing
   information loops:

   ORIGINATOR_ID

   ORIGINATOR_ID is a new optional, non-transitive BGP attribute of Type
   code 9.  This attribute is 4 bytes long and it will be created by an
   RR in reflecting a route.  This attribute will carry the BGP
   Identifier of the originator of the route in the local AS.  A BGP
   speaker SHOULD NOT create an ORIGINATOR_ID attribute if one already
   exists.  A router that recognizes the ORIGINATOR_ID attribute SHOULD
   ignore a route received with its BGP Identifier as the ORIGINATOR_ID.








Bates, et al.               Standards Track                     [Page 6]

RFC 4456                  BGP Route Reflection                April 2006


   CLUSTER_LIST

   CLUSTER_LIST is a new, optional, non-transitive BGP attribute of Type
   code 10.  It is a sequence of CLUSTER_ID values representing the
   reflection path that the route has passed.

   When an RR reflects a route, it MUST prepend the local CLUSTER_ID to
   the CLUSTER_LIST.  If the CLUSTER_LIST is empty, it MUST create a new
   one.  Using this attribute an RR can identify if the routing
   information has looped back to the same cluster due to
   misconfiguration.  If the local CLUSTER_ID is found in the
   CLUSTER_LIST, the advertisement received SHOULD be ignored.

9.  Impact on Route Selection



   The BGP Decision Process Tie Breaking rules (Sect.  9.1.2.2, [1]) are
   modified as follows:

      If a route carries the ORIGINATOR_ID attribute, then in Step f)
      the ORIGINATOR_ID SHOULD be treated as the BGP Identifier of the
      BGP speaker that has advertised the route.

      In addition, the following rule SHOULD be inserted between Steps
      f) and g): a BGP Speaker SHOULD prefer a route with the shorter
      CLUSTER_LIST length.  The CLUSTER_LIST length is zero if a route
      does not carry the CLUSTER_LIST attribute.

10.  Implementation Considerations



   Care should be taken to make sure that none of the BGP path
   attributes defined above can be modified through configuration when
   exchanging internal routing information between RRs and Clients and
   Non-Clients.  Their modification could potentially result in routing
   loops.

   In addition, when a RR reflects a route, it SHOULD NOT modify the
   following path attributes: NEXT_HOP, AS_PATH, LOCAL_PREF, and MED.
   Their modification could potentially result in routing loops.

11.  Configuration and Deployment Considerations



   The BGP protocol provides no way for a Client to identify itself
   dynamically as a Client of an RR.  The simplest way to achieve this
   is by manual configuration.

   One of the key component of the route reflection approach in
   addressing the scaling issue is that the RR summarizes routing
   information and only reflects its best path.



Bates, et al.               Standards Track                     [Page 7]

RFC 4456                  BGP Route Reflection                April 2006


   Both Multi-Exit Discriminators (MEDs) and Interior Gateway Protocol
   (IGP) metrics may impact the BGP route selection.  Because MEDs are
   not always comparable and the IGP metric may differ for each router,
   with certain route reflection topologies the route reflection
   approach may not yield the same route selection result as that of the
   full IBGP mesh approach.  A way to make route selection the same as
   it would be with the full IBGP mesh approach is to make sure that
   route reflectors are never forced to perform the BGP route selection
   based on IGP metrics that are significantly different from the IGP
   metrics of their clients, or based on incomparable MEDs.  The former
   can be achieved by configuring the intra-cluster IGP metrics to be
   better than the inter-cluster IGP metrics, and maintaining full mesh
   within the cluster.  The latter can be achieved by

      o  setting the local preference of a route at the border router to
         reflect the MED values, or

      o  making sure the AS-path lengths from different ASes are
         different when the AS-path length is used as a route selection
         criteria, or

      o  configuring community-based policies to influence the route
         selection.

   One could argue though that the latter requirement is overly
   restrictive, and perhaps impractical in some cases.  One could
   further argue that as long as there are no routing loops, there are
   no compelling reasons to force route selection with route reflectors
   to be the same as it would be with the full IBGP mesh approach.

   To prevent routing loops and maintain consistent routing view, it is
   essential that the network topology be carefully considered in
   designing a route reflection topology.  In general, the route
   reflection topology should be congruent with the network topology
   when there exist multiple paths for a prefix.  One commonly used
   approach is the reflection based on Point of Presence (POP), in which
   each POP maintains its own route reflectors serving clients in the
   POP, and all route reflectors are fully meshed.  In addition, clients
   of the reflectors in each POP are often fully meshed for the purpose
   of optimal intra-POP routing, and the intra-POP IGP metrics are
   configured to be better than the inter-POP IGP metrics.

12.  Security Considerations



   This extension to BGP does not change the underlying security issues
   inherent in the existing IBGP [1, 5].





Bates, et al.               Standards Track                     [Page 8]

RFC 4456                  BGP Route Reflection                April 2006


13.  Acknowledgements



   The authors would like to thank Dennis Ferguson, John Scudder, Paul
   Traina, and Tony Li for the many discussions resulting in this work.
   This idea was developed from an earlier discussion between Tony Li
   and Dimitri Haskin.

   In addition, the authors would like to acknowledge valuable review
   and suggestions from Yakov Rekhter on this document, and helpful
   comments from Tony Li, Rohit Dube, John Scudder, and Bruce Cole.

14.  References



14.1.  Normative References



   [1]  Rekhter, Y., Li, T., and S. Hares, "A Border Gateway Protocol 4
        (BGP-4)", RFC 4271, January 2006.

14.2.  Informative References



   [2]  Savola, P., "Reclassification of RFC 1863 to Historic", RFC
        4223, October 2005.

   [3]  Traina, P., McPherson, D., and J. Scudder, "Autonomous System
        Confederations for BGP", RFC 3065, February 2001.

   [4]  Bates, T. and R. Chandra, "BGP Route Reflection An alternative
        to full mesh IBGP", RFC 1966, June 1996.

   [5]  Heffernan, A., "Protection of BGP Sessions via the TCP MD5
        Signature Option", RFC 2385, August 1998.

   [6]  Bates, T., Chandra, R., and E. Chen, "BGP Route Reflection - An
        Alternative to Full Mesh IBGP", RFC 2796, April 2000.

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














Bates, et al.               Standards Track                     [Page 9]

RFC 4456                  BGP Route Reflection                April 2006


Appendix A: Comparison with RFC 2796



   The impact on route selection is added.

   The pictorial description of the encoding of the CLUSTER_LIST
   attribute is removed as the description is redundant to the BGP
   specification, and the attribute length field is inadvertently
   described as one octet.

Appendix B: Comparison with RFC 1966



   All the changes listed in Appendix A, plus the following.

   Several terminologies related to route reflection are clarified, and
   the reference to EBGP routes/peers are removed.

   The handling of a routing information loop (due to route reflection)
   by a receiver is clarified and made more consistent.

   The addition of a CLUSTER_ID to the CLUSTER_LIST has been changed
   from "append" to "prepend" to reflect the deployed code.

   The section on "Configuration and Deployment Considerations" has been
   expanded to address several operational issues.



























Bates, et al.               Standards Track                    [Page 10]

RFC 4456                  BGP Route Reflection                April 2006


Authors' Addresses



   Tony Bates
   Cisco Systems, Inc.
   170 West Tasman Drive
   San Jose, CA 95134

   EMail: tbates@cisco.com


   Ravi Chandra
   Sonoa Systems, Inc.
   3255-7 Scott Blvd.
   Santa Clara, CA 95054

   EMail: rchandra@sonoasystems.com


   Enke Chen
   Cisco Systems, Inc.
   170 West Tasman Drive
   San Jose, CA 95134

   EMail: enkechen@cisco.com



























Bates, et al.               Standards Track                    [Page 11]

RFC 4456                  BGP Route Reflection                April 2006


Full Copyright Statement



   Copyright (C) The Internet Society (2006).

   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-ipr@ietf.org.

Acknowledgement



   Funding for the RFC Editor function is provided by the IETF
   Administrative Support Activity (IASA).







Bates, et al.               Standards Track                    [Page 12]