Home Articles FAQs XREF Games Software Instant Books BBS About FOLDOC RFCs Feedback Sitemap
irt.Org

Request For Comments - RFC1774

You are here: irt.org | RFCs | RFC1774 [ previous next ]






Network Working Group                                  P. Traina, Editor
Request for Comments: 1774                                 cisco Systems
Category: Informational                                       March 1995

                        BGP-4 Protocol Analysis

Status of this Memo

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

Introduction

   The purpose of this report is to document how the requirements for
   advancing a routing protocol to Draft Standard have been satisfied by
   the Border Gateway Protocol version 4 (BGP-4). This report summarizes
   the key features of BGP, and analyzes the protocol with respect to
   scaling and performance. This is the first of two reports on the BGP
   protocol.

   BGP-4 is an inter-autonomous system routing protocol designed for
   TCP/IP internets.  Version 1 of the BGP protocol was published in RFC
   1105. Since then BGP versions 2, 3, and 4 have been developed.
   Version 2 was documented in RFC 1163. Version 3 is documented in
   RFC1267.  The changes between versions are explained in Appendix 2 of
   [1].

   Possible applications of BGP in the Internet are documented in [2].

   Please send comments to iwg@ans.net.

Key features and algorithms of the BGP-4 protocol.

   This section summarizes the key features and algorithms of the BGP
   protocol. BGP is an inter-autonomous system routing protocol; it is
   designed to be used between multiple autonomous systems. BGP assumes
   that routing within an autonomous system is done by an intra-
   autonomous system routing protocol. BGP does not make any assumptions
   about intra-autonomous system routing protocols employed by the
   various autonomous systems.  Specifically, BGP does not require all
   autonomous systems to run the same intra-autonomous system routing
   protocol.

   BGP is a real inter-autonomous system routing protocol. It imposes no
   constraints on the underlying Internet topology. The information
   exchanged via BGP is sufficient to construct a graph of autonomous
   systems connectivity from which routing loops may be pruned and some



Traina                                                          [Page 1]



RFC 1774                BGP-4 Protocol Analysis               March 1995


   routing policy decisions at the autonomous system level may be
   enforced.

   The key features of the protocol are the notion of path attributes
   and aggregation of network layer reachability information (NLRI).

   Path attributes provide BGP with flexibility and expandability. Path
   attributes are partitioned into well-known and optional. The
   provision for optional attributes allows experimentation that may
   involve a group of BGP routers without affecting the rest of the
   Internet.  New optional attributes can be added to the protocol in
   much the same fashion as new options are added to the Telnet
   protocol, for instance.

   One of the most important path attributes is the AS-PATH. AS
   reachability information traverses the Internet, this information is
   augmented by the list of autonomous systems that have been traversed
   thus far, forming the AS-PATH.  The AS-PATH allows straightforward
   suppression of the looping of routing information. In addition, the
   AS-PATH serves as a powerful and versatile mechanism for policy-based
   routing.

   BGP-4 enhances the AS-PATH attribute to include sets of autonomous
   systems as well as lists.  This extended format allows generated
   aggregate routes to carry path information from the more specific
   routes used to generate the aggregate.

   BGP uses an algorithm that cannot be classified as either a pure
   distance vector, or a pure link state. Carrying a complete AS path in
   the AS-PATH attribute allows to reconstruct large portions of the
   overall topology. That makes it similar to the link state algorithms.
   Exchanging only the currently used routes between the peers makes it
   similar to the distance vector algorithms.

   To conserve bandwidth and processing power, BGP uses incremental
   updates, where after the initial exchange of complete routing
   information, a pair of BGP routers exchanges only changes (deltas) to
   that information. Technique of incremental updates requires reliable
   transport between a pair of BGP routers. To achieve this
   functionality BGP uses TCP as its transport.

   In addition to incremental updates, BGP-4 has added the concept of
   route aggregation so that information about groups of networks may
   represented as a single entity.

   BGP is a self-contained protocol. That is, it specifies how routing
   information is exchanged both between BGP speakers in different
   autonomous systems, and between BGP speakers within a single



Traina                                                          [Page 2]



RFC 1774                BGP-4 Protocol Analysis               March 1995


   autonomous system.

   To allow graceful coexistence with EGP and OSPF, BGP provides support
   for carrying both EGP and OSPF derived exterior routes BGP also
   allows to carry statically defined exterior routes or routes derived
   by other IGP information.

BGP performance characteristics and scalability

   In this section we'll try to answer the question of how much link
   bandwidth, router memory and router CPU cycles does the BGP protocol
   consume under normal conditions.  We'll also address the scalability
   of BGP, and look at some of its limits.

   BGP does not require all the routers within an autonomous system to
   participate in the BGP protocol. Only the border routers that provide
   connectivity between the local autonomous system and its adjacent
   autonomous systems participate in BGP.  Constraining the set of
   participants is just one way of addressing the scaling issue.

Link bandwidth and CPU utilization

   Immediately after the initial BGP connection setup, the peers
   exchange complete set of routing information. If we denote the total
   number of routes in the Internet by N, the mean AS distance of the
   Internet by M (distance at the level of an autonomous system,
   expressed in terms of the number of autonomous systems), the total
   number of autonomous systems in the Internet by A, and assume that
   the networks are uniformly distributed among the autonomous systems,
   then the worst case amount of bandwidth consumed during the initial
   exchange between a pair of BGP speakers is

                    MR = O(N + M * A)

   The following table illustrates typical amount of bandwidth consumed
   during the initial exchange between a pair of BGP speakers based on
   the above assumptions (ignoring bandwidth consumed by the BGP
   Header).

   # NLRI       Mean AS Distance       # AS's    Bandwidth
   ----------   ----------------       ------    ---------
   10,000       15                     300       49,000 bytes
   20,000       8                      400       86,000 bytes *
   40,000       15                     400       172,000 bytes
   100,000      20                     3,000     520,000 bytes

   * the actual "size" of the Internet at the the time of this
     document's publication



Traina                                                          [Page 3]



RFC 1774                BGP-4 Protocol Analysis               March 1995


   Note that most of the bandwidth is consumed by the exchange of the
   Network Layer Reachability Information (NLRI).

   BGP-4 was created specifically to reduce the amount of NLRI entries
   carried and exchanged by border routers.  BGP-4, along with CIDR [4]
   has introduced the concept of the "Supernet" which describes a
   power-of-two aggregation of more than one class-based network.

   Due to the advantages of advertising a few large aggregate blocks
   instead of many smaller class-based individual networks, it is
   difficult to estimate the actual reduction in bandwidth and
   processing that BGP-4 has provided over BGP3.  If we simply enumerate
   all aggregate blocks into their individual class-based networks, we
   would not take into account "dead" space that has been reserved for
   future expansion.  The best metric for determining the success of
   BGP-4's aggregation is to sample the number NLRI entries in the
   globally connected Internet today and compare it to projected growth
   rates before BGP-4 was deployed.

   In January of 1994, router carrying a full routing load for the
   globally connected Internet had approximately 19,000 network entries
   (this number is not exact due to local policy variations).  The BGP
   deployment working group estimated that the growth rate at that time
   was over 1000 new networks per month and increasing.  Since the
   widespread deployment of BGP-4, the growth rate has dropped
   significantly and a sample done at the end of November 1994 showed
   approximately 21,000 entries present,  as opposed to the expected
   30,000.

   CPU cycles consumed by BGP depends only on the stability of the
   Internet. If the Internet is stable, then the only link bandwidth and
   router CPU cycles consumed by BGP are due to the exchange of the BGP
   KEEPALIVE messages. The KEEPALIVE messages are exchanged only between
   peers. The suggested frequency of the exchange is 30 seconds. The
   KEEPALIVE messages are quite short (19 octets), and require virtually
   no processing.  Therefore, the bandwidth consumed by the KEEPALIVE
   messages is about 5 bits/sec.  Operational experience confirms that
   the overhead (in terms of bandwidth and CPU) associated with the
   KEEPALIVE messages should be viewed as negligible.  If the Internet
   is unstable, then only the changes to the reachability information
   (that are caused by the instabilities) are passed between routers
   (via the UPDATE messages). If we denote the number of routing changes
   per second by C, then in the worst case the amount of bandwidth
   consumed by the BGP can be expressed as O(C * M). The greatest
   overhead per UPDATE message occurs when each UPDATE message contains
   only a single network. It should be pointed out that in practice
   routing changes exhibit strong locality with respect to the AS path.
   That is routes that change are likely to have common AS path. In this



Traina                                                          [Page 4]



RFC 1774                BGP-4 Protocol Analysis               March 1995


   case multiple networks can be grouped into a single UPDATE message,
   thus significantly reducing the amount of bandwidth required (see
   also Appendix 6.1 of [1]).

   Since in the steady state the link bandwidth and router CPU cycles
   consumed by the BGP protocol are dependent only on the stability of
   the Internet, but are completely independent on the number of
   networks that compose the Internet, it follows that BGP should have
   no scaling problems in the areas of link bandwidth and router CPU
   utilization, as the Internet grows, provided that the overall
   stability of the inter-AS connectivity (connectivity between ASs) of
   the Internet can be controlled. Stability issue could be addressed by
   introducing some form of dampening (e.g., hold downs).  Due to the
   nature of BGP, such dampening should be viewed as a local to an
   autonomous system matter (see also Appendix 6.3 of [1]). It is
   important to point out, that regardless of BGP, one should not
   underestimate the significance of the stability in the Internet.

   Growth of the Internet has made the stability issue one of the most
   crucial ones. It is important to realize that BGP, by itself, does
   not introduce any instabilities in the Internet. Current observations
   in the NSFNET show that the instabilities are largely due to the
   ill-behaved routing within the autonomous systems that compose the
   Internet.  Therefore, while providing BGP with mechanisms to address
   the stability issue, we feel that the right way to handle the issue
   is to address it at the root of the problem, and to come up with
   intra-autonomous routing schemes that exhibit reasonable stability.

   It also may be instructive to compare bandwidth and CPU requirements
   of BGP with EGP. While with BGP the complete information is exchanged
   only at the connection establishment time, with EGP the complete
   information is exchanged periodically (usually every 3 minutes). Note
   that both for BGP and for EGP the amount of information exchanged is
   roughly on the order of the networks reachable via a peer that sends
   the information (see also Section 5.2). Therefore, even if one
   assumes extreme instabilities of BGP, its worst case behavior will be
   the same as the steady state behavior of EGP.

   Operational experience with BGP showed that the incremental updates
   approach employed by BGP presents an enormous improvement both in the
   area of bandwidth and in the CPU utilization, as compared with
   complete periodic updates used by EGP (see also presentation by
   Dennis Ferguson at the Twentieth IETF, March 11-15, 1991, St.Louis).








Traina                                                          [Page 5]



RFC 1774                BGP-4 Protocol Analysis               March 1995


Memory requirements

   To quantify the worst case memory requirements for BGP, denote the
   total number of networks in the Internet by N, the mean AS distance
   of the Internet by M (distance at the level of an autonomous system,
   expressed in terms of the number of autonomous systems), the total
   number of autonomous systems in the Internet by A, and the total
   number of BGP speakers that a system is peering with by K (note that
   K will usually be dominated by the total number of the BGP speakers
   within a single autonomous system). Then the worst case memory
   requirements (MR) can be expressed as

                    MR = O((N + M * A) * K)

   In the current NSFNET Backbone (N = 2110, A = 59, and M = 5) if each
   network is stored as 4 octets, and each autonomous system is stored
   as 2 octets then the overhead of storing the AS path information (in
   addition to the full complement of exterior routes) is less than 7
   percent of the total memory usage.

   It is interesting to point out, that prior to the introduction of BGP
   in the NSFNET Backbone, memory requirements on the NSFNET Backbone
   routers running EGP were on the order of O(N * K). Therefore, the
   extra overhead in memory incurred by the NSFNET routers after the
   introduction of BGP is less than 7 percent.

   Since a mean AS distance grows very slowly with the total number of
   networks (there are about 60 autonomous systems, well over 2,000
   networks known in the NSFNET backbone routers, and the mean AS
   distance of the current Internet is well below 5), for all practical
   purposes the worst case router memory requirements are on the order
   of the total number of networks in the Internet times the number of
   peers the local system is peering with. We expect that the total
   number of networks in the Internet will grow much faster than the
   average number of peers per router. Therefore, scaling with respect
   to the memory requirements is going to be heavily dominated by the
   factor that is linearly proportional to the total number of networks
   in the Internet.

   The following table illustrates typical memory requirements of a
   router running BGP. It is assumed that each network is encoded as 4
   bytes, each AS is encoded as 2 bytes, and each networks is reachable
   via some fraction of all of the peers (# BGP peers/per net).








Traina                                                          [Page 6]



RFC 1774                BGP-4 Protocol Analysis               March 1995


   # Networks  Mean AS Distance # AS's # BGP peers/per net Memory Req
   ----------  ---------------- ------ ------------------- ----------
   2,100       5                59     3                   27,000
   4,000       10               100    6                   108,000
   10,000      15               300    10                  490,000
   100,000     20               3,000  20                  1,040,000

   To put memory requirements of BGP in a proper perspective, let's try
   to put aside for a moment the issue of what information is used to
   construct the forwarding tables in a router, and just focus on the
   forwarding tables themselves. In this case one might ask about the
   limits on these tables.  For instance, given that right now the
   forwarding tables in the NSFNET Backbone routers carry well over
   20,000 entries, one might ask whether it would be possible to have a
   functional router with a table that will have 200,000 entries.
   Clearly the answer to this question is completely independent of BGP.
   On the other hand the answer to the original questions (that was
   asked with respect to BGP) is directly related to the latter
   question. Very interesting comments were given by Paul Tsuchiya in
   his review of BGP in March of 1990 (as part of the BGP review
   committee appointed by Bob Hinden).  In the review he said that, "BGP
   does not scale well.  This is not really the fault of BGP. It is the
   fault of the flat IP address space.  Given the flat IP address space,
   any routing protocol must carry network numbers in its updates." With
   the introduction of CIDR [4] and BGP-4,  we have attempted to reduce
   this limitation.  Unfortunately,  we cannot erase history nor can
   BGP-4 solve the problems inherent with inefficient assignment of
   future address blocks.

   To reiterate, BGP limits with respect to the memory requirements are
   directly related to the underlying Internet Protocol (IP), and
   specifically the addressing scheme employed by IP. BGP would provide
   much better scaling in environments with more flexible addressing
   schemes.  It should be pointed out that with only very minor
   additions BGP was extended to support hierarchies of autonomous
   system [8]. Such hierarchies, combined with an addressing scheme that
   would allow more flexible address aggregation capabilities, can be
   utilized by BGP-like protocols, thus providing practically unlimited
   scaling capabilities.

Applicability of BGP

   In this section we'll try to answer the question of what environment
   is BGP well suited, and for what is it not suitable?  Partially this
   question is answered in the Section 2 of [1], where the document
   states the following:





Traina                                                          [Page 7]



RFC 1774                BGP-4 Protocol Analysis               March 1995


      "To characterize the set of policy decisions that can be enforced
      using BGP, one must focus on the rule that an AS advertises to its
      neighbor ASs only those routes that it itself uses.  This rule
      reflects the "hop-by-hop" routing paradigm generally used
      throughout the current Internet.  Note that some policies cannot
      be supported by the "hop-by-hop" routing paradigm and thus require
      techniques such as source routing to enforce.  For example, BGP
      does not enable one AS to send traffic to a neighbor AS intending
      that the traffic take a different route from that taken by traffic
      originating in the neighbor AS.  On the other hand, BGP can
      support any policy conforming to the "hop-by-hop" routing
      paradigm.  Since the current Internet uses only the "hop-by-hop"
      routing paradigm and since BGP can support any policy that
      conforms to that paradigm, BGP is highly applicable as an inter-AS
      routing protocol for the current Internet."

   While BGP is well suitable for the current Internet, it is also
   almost a necessity for the current Internet as well.  Operational
   experience with EGP showed that it is highly inadequate for the
   current Internet.  Topological restrictions imposed by EGP are
   unjustifiable from the technical point of view, and unenforceable
   from the practical point of view.  Inability of EGP to efficiently
   handle information exchange between peers is a cause of severe
   routing instabilities in the operational Internet. Finally,
   information provided by BGP is well suitable for enforcing a variety
   of routing policies.

   Rather than trying to predict the future, and overload BGP with a
   variety of functions that may (or may not) be needed, the designers
   of BGP took a different approach. The protocol contains only the
   functionality that is essential, while at the same time provides
   flexible mechanisms within the protocol itself that allow to expand
   its functionality.  Since BGP was designed with flexibility and
   expandability in mind, we think it should be able to address new or
   evolving requirements with relative ease. The existence proof of this
   statement may be found in the way how new features (like repairing a
   partitioned autonomous system with BGP) are already introduced in the
   protocol.

   To summarize, BGP is well suitable as an inter-autonomous system
   routing protocol for the current Internet that is based on IP (RFC
   791) as the Internet Protocol and "hop-by-hop" routing paradigm. It
   is hard to speculate whether BGP will be suitable for other
   environments where internetting is done by other than IP protocols,
   or where the routing paradigm will be different.






Traina                                                          [Page 8]



RFC 1774                BGP-4 Protocol Analysis               March 1995


Security Considerations

   Security issues are not discussed in this memo.

Acknowledgments

   The BGP-4 protocol has been developed by the IDR/BGP Working Group of
   the Internet Engineering Task Force.  I would like to express thanks
   to Yakov Rekhter for providing RFC 1265.  This document is only a
   minor update to the original text. I'd also like to explicitly thank
   Yakov Rekhter and Tony Li for their review of this document as well
   as their constructive and valuable comments.

Editor's Address

   Paul Traina
   cisco Systems, Inc.
   170 W. Tasman Dr.
   San Jose, CA 95134

   EMail: pst@cisco.com






























Traina                                                          [Page 9]



RFC 1774                BGP-4 Protocol Analysis               March 1995


References

   [1] Rekhter, Y., and T., Li, "A Border Gateway Protocol 4 (BGP-4)",
       RFC 1771, T.J. Watson Research Center, IBM Corp., cisco Systems,
       March 1995.

   [2] Rekhter, Y., and P. Gross, Editors, "Application of the Border
       Gateway Protocol in the Internet", RFC 1772, T.J. Watson Research
       Center, IBM Corp., MCI, March 1995.

   [3] Willis, S., Burruss, J., and J. Chu, "Definitions of Managed
       Objects for the Fourth Version of the Border Gateway Protocol
       (BGP-4) using SMIv2", RFC 1657, Wellfleet Communications Inc.,
       IBM Corp., July 1994.

   [4] Fuller V., Li. T., Yu J., and K. Varadhan, "Classless Inter-
       Domain Routing (CIDR): an Address Assignment and Aggregation
       Strategy", RFC 1519, BARRNet, cisco, MERIT, OARnet, September
       1993.

   [6] Moy J., "Open Shortest Path First Routing Protocol (Version 2)",
       RFC 1257, Proteon, August 1991.

   [7] Varadhan, K., Hares S., and Y. Rekhter, "BGP4/IDRP for IP---OSPF
       Interaction", Work in Progress.

   [8] ISO/IEC 10747, Kunzinger, C., Editor, "Inter-Domain Routing
       Protocol", October 1993.























Traina                                                         [Page 10]



©2018 Martin Webb