Network Working Group Y. Rekhter Request for Comments: 1771 T.J. Watson Research Center, IBM Corp. Obsoletes: 1654 T. Li Category: Standards Track cisco Systems Editors March 1995
A Border Gateway Protocol 4 (BGP-4)
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.
This document, together with its companion document, "Application of the Border Gateway Protocol in the Internet", define an inter- autonomous system routing protocol for the Internet.
This document was originally published as RFC 1267 in October 1991, jointly authored by Kirk Lougheed (cisco Systems) and Yakov Rekhter (IBM).
We would like to express our thanks to Guy Almes (ANS), Len Bosack (cisco Systems), and Jeffrey C. Honig (Cornell University) for their contributions to the earlier version of this document.
We like to explicitly thank Bob Braden (ISI) for the review of the earlier version of this document as well as his constructive and valuable comments.
We would also like to thank Bob Hinden, Director for Routing of the Internet Engineering Steering Group, and the team of reviewers he assembled to review the previous version (BGP-2) of this document. This team, consisting of Deborah Estrin, Milo Medin, John Moy, Radia Perlman, Martha Steenstrup, Mike St. Johns, and Paul Tsuchiya, acted with a strong combination of toughness, professionalism, and courtesy.
This updated version of the document is the product of the IETF IDR Working Group with Yakov Rekhter and Tony Li as editors. Certain sections of the document borrowed heavily from IDRP , which is the OSI counterpart of BGP. For this credit should be given to the ANSI X3S3.3 group chaired by Lyman Chapin (BBN) and to Charles Kunzinger (IBM Corp.) who was the IDRP editor within that group. We would also like to thank Mike Craren (Proteon, Inc.), Dimitry Haskin (Bay Networks, Inc.), John Krawczyk (Bay Networks, Inc.), and Paul Traina (cisco Systems) for their insightful comments.
We would like to specially acknowledge numerous contributions by Dennis Ferguson (MCI).
The work of Yakov Rekhter was supported in part by the National Science Foundation under Grant Number NCR-9219216.
The Border Gateway Protocol (BGP) is an inter-Autonomous System routing protocol. It is built on experience gained with EGP as defined in RFC 904  and EGP usage in the NSFNET Backbone as described in RFC 1092  and RFC 1093 .
The primary function of a BGP speaking system is to exchange network reachability information with other BGP systems. This network reachability information includes information on the list of Autonomous Systems (ASs) that reachability information traverses. This information is sufficient to construct a graph of AS connectivity from which routing loops may be pruned and some policy decisions at the AS level may be enforced.
BGP-4 provides a new set of mechanisms for supporting classless interdomain routing. These mechanisms include support for advertising an IP prefix and eliminates the concept of network "class" within BGP. BGP-4 also introduces mechanisms which allow aggregation of routes, including aggregation of AS paths. These changes provide support for the proposed supernetting scheme [8, 9].
To characterize the set of policy decisions that can be enforced using BGP, one must focus on the rule that a BGP speaker advertise to its peers (other BGP speakers which it communicates with) in neighboring 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 neighboring AS intending that the traffic take a different route from that taken by traffic originating in the neighboring 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.
A more complete discussion of what policies can and cannot be enforced with BGP is outside the scope of this document (but refer to the companion document discussing BGP usage ).
BGP runs over a reliable transport protocol. This eliminates the need to implement explicit update fragmentation, retransmission, acknowledgement, and sequencing. Any authentication scheme used by the transport protocol may be used in addition to BGP's own authentication mechanisms. The error notification mechanism used in BGP assumes that the transport protocol supports a "graceful" close, i.e., that all outstanding data will be delivered before the connection is closed.
BGP uses TCP  as its transport protocol. TCP meets BGP's transport requirements and is present in virtually all commercial routers and hosts. In the following descriptions the phrase "transport protocol connection" can be understood to refer to a TCP connection. BGP uses TCP port 179 for establishing its connections.
This document uses the term `Autonomous System' (AS) throughout. The classic definition of an Autonomous System is a set of routers under a single technical administration, using an interior gateway protocol and common metrics to route packets within the AS, and using an exterior gateway protocol to route packets to other ASs. Since this classic definition was developed, it has become common for a single AS to use several interior gateway protocols and sometimes several sets of metrics within an AS. The use of the term Autonomous System here stresses the fact that, even when multiple IGPs and metrics are used, the administration of an AS appears to other ASs to have a single coherent interior routing plan and presents a consistent picture of what destinations are reachable through it.
The planned use of BGP in the Internet environment, including such issues as topology, the interaction between BGP and IGPs, and the enforcement of routing policy rules is presented in a companion document . This document is the first of a series of documents planned to explore various aspects of BGP application. Please send comments to the BGP mailing list (firstname.lastname@example.org)
3. Summary of Operation
Two systems form a transport protocol connection between one another. They exchange messages to open and confirm the connection parameters. The initial data flow is the entire BGP routing table. Incremental updates are sent as the routing tables change. BGP does not require periodic refresh of the entire BGP routing table. Therefore, a BGP speaker must retain the current version of the entire BGP routing tables of all of its peers for the duration of the connection. KeepAlive messages are sent periodically to ensure the liveness of the connection. Notification messages are sent in response to errors or special conditions. If a connection encounters an error condition, a notification message is sent and the connection is closed.
The hosts executing the Border Gateway Protocol need not be routers. A non-routing host could exchange routing information with routers via EGP or even an interior routing protocol. That non-routing host could then use BGP to exchange routing information with a border router in another Autonomous System. The implications and applications of this architecture are for further study.
If a particular AS has multiple BGP speakers and is providing transit service for other ASs, then care must be taken to ensure a consistent view of routing within the AS. A consistent view of the interior routes of the AS is provided by the interior routing protocol. A consistent view of the routes exterior to the AS can be provided by having all BGP speakers within the AS maintain direct BGP connections with each other. Using a common set of policies, the BGP speakers arrive at an agreement as to which border routers will serve as exit/entry points for particular destinations outside the AS. This information is communicated to the AS's internal routers, possibly via the interior routing protocol. Care must be taken to ensure that the interior routers have all been updated with transit information before the BGP speakers announce to other ASs that transit service is being provided.
Connections between BGP speakers of different ASs are referred to as "external" links. BGP connections between BGP speakers within the same AS are referred to as "internal" links. Similarly, a peer in a different AS is referred to as an external peer, while a peer in the same AS may be described as an internal peer.
3.1 Routes: Advertisement and Storage
For purposes of this protocol a route is defined as a unit of information that pairs a destination with the attributes of a path to that destination:
If a BGP speaker chooses to advertise the route, it may add to or modify the path attributes of the route before advertising it to a peer.
BGP provides mechanisms by which a BGP speaker can inform its peer that a previously advertised route is no longer available for use. There are three methods by which a given BGP speaker can indicate that a route has been withdrawn from service:
3.2 Routing Information Bases
The Routing Information Base (RIB) within a BGP speaker consists of three distinct parts:
In summary, the Adj-RIBs-In contain unprocessed routing information that has been advertised to the local BGP speaker by its peers; the Loc-RIB contains the routes that have been selected by the local BGP speaker's Decision Process; and the Adj-RIBs-Out organize the routes for advertisement to specific peers by means of the local speaker's UPDATE messages.
Although the conceptual model distinguishes between Adj-RIBs-In, Loc-RIB, and Adj-RIBs-Out, this neither implies nor requires that an implementation must maintain three separate copies of the routing information. The choice of implementation (for example, 3 copies of the information vs 1 copy with pointers) is not constrained by the protocol.
4. Message Formats
This section describes message formats used by BGP.
Messages are sent over a reliable transport protocol connection. A message is processed only after it is entirely received. The maximum message size is 4096 octets. All implementations are required to support this maximum message size. The smallest message that may be sent consists of a BGP header without a data portion, or 19 octets.
4.1 Message Header Format
Each message has a fixed-size header. There may or may not be a data portion following the header, depending on the message type. The layout of these fields is shown below:
4.2 OPEN Message Format
After a transport protocol connection is established, the first message sent by each side is an OPEN message. If the OPEN message is acceptable, a KEEPALIVE message confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, KEEPALIVE, and NOTIFICATION messages may be exchanged.
In addition to the fixed-size BGP header, the OPEN message contains the following fields:
My Autonomous System:
Optional Parameters Length:
This document defines the following Optional Parameters:
a) Authentication Information (Parameter Type 1):
0 1 2 3 4 5 6 7 8
Note that a separate authentication mechanism may be used in establishing the transport level connection.
The minimum length of the OPEN message is 29 octets (including message header).
4.3 UPDATE Message Format
UPDATE messages are used to transfer routing information between BGP peers. The information in the UPDATE packet can be used to construct a graph describing the relationships of the various Autonomous Systems. By applying rules to be discussed, routing information loops and some other anomalies may be detected and removed from inter-AS routing.
An UPDATE message is used to advertise a single feasible route to a peer, or to withdraw multiple unfeasible routes from service (see 3.1). An UPDATE message may simultaneously advertise a feasible route and withdraw multiple unfeasible routes from service. The UPDATE message always includes the fixed-size BGP header, and can optionally include the other fields as shown below:
Network Layer Reachability Information:
The minimum length of the UPDATE message is 23 octets -- 19 octets for the fixed header + 2 octets for the Unfeasible Routes Length + 2 octets for the Total Path Attribute Length (the value of Unfeasible Routes Length is 0 and the value of Total Path Attribute Length is 0).
An UPDATE message can advertise at most one route, which may be described by several path attributes. All path attributes contained in a given UPDATE messages apply to the destinations carried in the Network Layer Reachability Information field of the UPDATE message.
An UPDATE message can list multiple routes to be withdrawn from service. Each such route is identified by its destination (expressed as an IP prefix), which unambiguously identifies the route in the context of the BGP speaker - BGP speaker connection to which it has been previously been advertised.
An UPDATE message may advertise only routes to be withdrawn from service, in which case it will not include path attributes or Network Layer Reachability Information. Conversely, it may advertise only a feasible route, in which case the WITHDRAWN ROUTES field need not be present.
4.4 KEEPALIVE Message Format
BGP does not use any transport protocol-based keep-alive mechanism to determine if peers are reachable. Instead, KEEPALIVE messages are exchanged between peers often enough as not to cause the Hold Timer to expire. A reasonable maximum time between KEEPALIVE messages would be one third of the Hold Time interval. KEEPALIVE messages MUST NOT be sent more frequently than one per second. An implementation MAY adjust the rate at which it sends KEEPALIVE messages as a function of the Hold Time interval.
If the negotiated Hold Time interval is zero, then periodic KEEPALIVE messages MUST NOT be sent.
KEEPALIVE message consists of only message header and has a length of 19 octets.
4.5 NOTIFICATION Message Format
A NOTIFICATION message is sent when an error condition is detected. The BGP connection is closed immediately after sending it.
In addition to the fixed-size BGP header, the NOTIFICATION message contains the following fields:
The minimum length of the NOTIFICATION message is 21 octets (including message header).
5. Path Attributes
This section discusses the path attributes of the UPDATE message.
Path attributes fall into four separate categories:
Well-known attributes must be recognized by all BGP implementations. Some of these attributes are mandatory and must be included in every UPDATE message. Others are discretionary and may or may not be sent in a particular UPDATE message.
All well-known attributes must be passed along (after proper updating, if necessary) to other BGP peers.
In addition to well-known attributes, each path may contain one or more optional attributes. It is not required or expected that all BGP implementations support all optional attributes. The handling of an unrecognized optional attribute is determined by the setting of the Transitive bit in the attribute flags octet. Paths with unrecognized transitive optional attributes should be accepted. If a path with unrecognized transitive optional attribute is accepted and passed along to other BGP peers, then the unrecognized transitive optional attribute of that path must be passed along with the path to other BGP peers with the Partial bit in the Attribute Flags octet set to 1. If a path with recognized transitive optional attribute is accepted and passed along to other BGP peers and the Partial bit in the Attribute Flags octet is set to 1 by some previous AS, it is not set back to 0 by the current AS. Unrecognized non-transitive optional attributes must be quietly ignored and not passed along to other BGP peers.
New transitive optional attributes may be attached to the path by the originator or by any other AS in the path. If they are not attached by the originator, the Partial bit in the Attribute Flags octet is set to 1. The rules for attaching new non-transitive optional attributes will depend on the nature of the specific attribute. The documentation of each new non-transitive optional attribute will be expected to include such rules. (The description of the MULTI_EXIT_DISC attribute gives an example.) All optional attributes (both transitive and non-transitive) may be updated (if appropriate) by ASs in the path.
The sender of an UPDATE message should order path attributes within the UPDATE message in ascending order of attribute type. The receiver of an UPDATE message must be prepared to handle path attributes within the UPDATE message that are out of order.
The same attribute cannot appear more than once within the Path Attributes field of a particular UPDATE message.
5.1 Path Attribute Usage
The usage of each BGP path attributes is described in the following clauses.
ORIGIN is a well-known mandatory attribute. The ORIGIN attribute shall be generated by the autonomous system that originates the associated routing information. It shall be included in the UPDATE messages of all BGP speakers that choose to propagate this information to other BGP speakers.
AS_PATH is a well-known mandatory attribute. This attribute identifies the autonomous systems through which routing information carried in this UPDATE message has passed. The components of this list can be AS_SETs or AS_SEQUENCEs.
When a BGP speaker propagates a route which it has learned from another BGP speaker's UPDATE message, it shall modify the route's AS_PATH attribute based on the location of the BGP speaker to which the route will be sent:
When a BGP speaker originates a route then:
The NEXT_HOP path attribute defines the IP address of the border router that should be used as the next hop to the destinations listed in the UPDATE message. If a border router belongs to the same AS as its peer, then the peer is an internal border router. Otherwise, it is an external border router. A BGP speaker can advertise any internal border router as the next hop provided that the interface associated with the IP address of this border router (as specified in the NEXT_HOP path attribute) shares a common subnet with both the local and remote BGP speakers. A BGP speaker can advertise any external border router as the next hop, provided that the IP address of this border router was learned from one of the BGP speaker's peers, and the interface associated with the IP address of this border router (as specified in the NEXT_HOP path attribute) shares a common subnet with the local and remote BGP speakers. A BGP speaker needs to be able to support disabling advertisement of external border routers.
A BGP speaker must never advertise an address of a peer to that peer as a NEXT_HOP, for a route that the speaker is originating. A BGP speaker must never install a route with itself as the next hop.
When a BGP speaker advertises the route to a BGP speaker located in its own autonomous system, the advertising speaker shall not modify the NEXT_HOP attribute associated with the route. When a BGP speaker receives the route via an internal link, it may forward packets to the NEXT_HOP address if the address contained in the attribute is on a common subnet with the local and remote BGP speakers.
The MULTI_EXIT_DISC attribute may be used on external (inter-AS) links to discriminate among multiple exit or entry points to the same neighboring AS. The value of the MULTI_EXIT_DISC attribute is a four octet unsigned number which is called a metric. All other factors being equal, the exit or entry point with lower metric should be preferred. If received over external links, the MULTI_EXIT_DISC attribute may be propagated over internal links to other BGP speakers within the same AS. The MULTI_EXIT_DISC attribute is never propagated to other BGP speakers in neighboring AS's.
LOCAL_PREF is a well-known discretionary attribute that shall be included in all UPDATE messages that a given BGP speaker sends to the other BGP speakers located in its own autonomous system. A BGP speaker shall calculate the degree of preference for each external route and include the degree of preference when advertising a route to its internal peers. The higher degree of preference should be preferred. A BGP speaker shall use the degree of preference learned via LOCAL_PREF in its decision process (see section 9.1.1).
A BGP speaker shall not include this attribute in UPDATE messages that it sends to BGP speakers located in a neighboring autonomous system. If it is contained in an UPDATE message that is received from a BGP speaker which is not located in the same autonomous system as the receiving speaker, then this attribute shall be ignored by the receiving speaker.
ATOMIC_AGGREGATE is a well-known discretionary attribute. If a BGP speaker, when presented with a set of overlapping routes from one of its peers (see 9.1.4), selects the less specific route without selecting the more specific one, then the local system shall attach the ATOMIC_AGGREGATE attribute to the route when propagating it to other BGP speakers (if that attribute is not already present in the received less specific route). A BGP speaker that receives a route with the ATOMIC_AGGREGATE attribute shall not remove the attribute from the route when propagating it to other speakers. A BGP speaker that receives a route with the ATOMIC_AGGREGATE attribute shall not make any NLRI of that route more specific (as defined in 9.1.4) when advertising this route to other BGP speakers. A BGP speaker that receives a route with the ATOMIC_AGGREGATE attribute needs to be cognizant of the fact that the actual path to destinations, as specified in the NLRI of the route, while having the loop-free property, may traverse ASs that are not listed in the AS_PATH attribute.
AGGREGATOR is an optional transitive attribute which may be included in updates which are formed by aggregation (see Section 126.96.36.199). A BGP speaker which performs route aggregation may add the AGGREGATOR attribute which shall contain its own AS number and IP address.
6. BGP Error Handling.
This section describes actions to be taken when errors are detected while processing BGP messages.
When any of the conditions described here are detected, a NOTIFICATION message with the indicated Error Code, Error Subcode, and Data fields is sent, and the BGP connection is closed. If no Error Subcode is specified, then a zero must be used.
The phrase "the BGP connection is closed" means that the transport protocol connection has been closed and that all resources for that BGP connection have been deallocated. Routing table entries associated with the remote peer are marked as invalid. The fact that the routes have become invalid is passed to other BGP peers before the routes are deleted from the system.
Unless specified explicitly, the Data field of the NOTIFICATION message that is sent to indicate an error is empty.
6.1 Message Header error handling.
All errors detected while processing the Message Header are indicated by sending the NOTIFICATION message with Error Code Message Header Error. The Error Subcode elaborates on the specific nature of the error.
The expected value of the Marker field of the message header is all ones if the message type is OPEN. The expected value of the Marker field for all other types of BGP messages determined based on the presence of the Authentication Information Optional Parameter in the BGP OPEN message and the actual authentication mechanism (if the Authentication Information in the BGP OPEN message is present). If the Marker field of the message header is not the expected one, then a synchronization error has occurred and the Error Subcode is set to Connection Not Synchronized.
If the Length field of the message header is less than 19 or greater than 4096, or if the Length field of an OPEN message is less than the minimum length of the OPEN message, or if the Length field of an UPDATE message is less than the minimum length of the UPDATE message, or if the Length field of a KEEPALIVE message is not equal to 19, or if the Length field of a NOTIFICATION message is less than the minimum length of the NOTIFICATION message, then the Error Subcode is set to Bad Message Length. The Data field contains the erroneous Length field.
If the Type field of the message header is not recognized, then the Error Subcode is set to Bad Message Type. The Data field contains the erroneous Type field.
6.2 OPEN message error handling.
All errors detected while processing the OPEN message are indicated by sending the NOTIFICATION message with Error Code OPEN Message Error. The Error Subcode elaborates on the specific nature of the error.
If the version number contained in the Version field of the received OPEN message is not supported, then the Error Subcode is set to Unsupported Version Number. The Data field is a 2-octet unsigned integer, which indicates the largest locally supported version number less than the version the remote BGP peer bid (as indicated in the received OPEN message).
If the Autonomous System field of the OPEN message is unacceptable, then the Error Subcode is set to Bad Peer AS. The determination of acceptable Autonomous System numbers is outside the scope of this protocol.
If the Hold Time field of the OPEN message is unacceptable, then the Error Subcode MUST be set to Unacceptable Hold Time. An implementation MUST reject Hold Time values of one or two seconds. An implementation MAY reject any proposed Hold Time. An implementation which accepts a Hold Time MUST use the negotiated value for the Hold Time.
If the BGP Identifier field of the OPEN message is syntactically incorrect, then the Error Subcode is set to Bad BGP Identifier. Syntactic correctness means that the BGP Identifier field represents a valid IP host address.
If one of the Optional Parameters in the OPEN message is not recognized, then the Error Subcode is set to Unsupported Optional Parameters.
If the OPEN message carries Authentication Information (as an Optional Parameter), then the corresponding authentication procedure is invoked. If the authentication procedure (based on Authentication Code and Authentication Data) fails, then the Error Subcode is set to Authentication Failure.
6.3 UPDATE message error handling.
All errors detected while processing the UPDATE message are indicated by sending the NOTIFICATION message with Error Code UPDATE Message Error. The error subcode elaborates on the specific nature of the error.
Error checking of an UPDATE message begins by examining the path attributes. If the Unfeasible Routes Length or Total Attribute Length is too large (i.e., if Unfeasible Routes Length + Total Attribute Length + 23 exceeds the message Length), then the Error Subcode is set to Malformed Attribute List.
If any recognized attribute has Attribute Flags that conflict with the Attribute Type Code, then the Error Subcode is set to Attribute Flags Error. The Data field contains the erroneous attribute (type, length and value).
If any recognized attribute has Attribute Length that conflicts with the expected length (based on the attribute type code), then the Error Subcode is set to Attribute Length Error. The Data field contains the erroneous attribute (type, length and value).
If any of the mandatory well-known attributes are not present, then the Error Subcode is set to Missing Well-known Attribute. The Data field contains the Attribute Type Code of the missing well-known attribute.
If any of the mandatory well-known attributes are not recognized, then the Error Subcode is set to Unrecognized Well-known Attribute. The Data field contains the unrecognized attribute (type, length and value).
If the ORIGIN attribute has an undefined value, then the Error Subcode is set to Invalid Origin Attribute. The Data field contains the unrecognized attribute (type, length and value).
If the NEXT_HOP attribute field is syntactically incorrect, then the Error Subcode is set to Invalid NEXT_HOP Attribute. The Data field contains the incorrect attribute (type, length and value). Syntactic correctness means that the NEXT_HOP attribute represents a valid IP host address. Semantic correctness applies only to the external BGP links. It means that the interface associated with the IP address, as specified in the NEXT_HOP attribute, shares a common subnet with the receiving BGP speaker and is not the IP address of the receiving BGP speaker. If the NEXT_HOP attribute is semantically incorrect, the error should be logged, and the the route should be ignored. In this case, no NOTIFICATION message should be sent.
The AS_PATH attribute is checked for syntactic correctness. If the path is syntactically incorrect, then the Error Subcode is set to Malformed AS_PATH.
If an optional attribute is recognized, then the value of this attribute is checked. If an error is detected, the attribute is discarded, and the Error Subcode is set to Optional Attribute Error. The Data field contains the attribute (type, length and value).
If any attribute appears more than once in the UPDATE message, then the Error Subcode is set to Malformed Attribute List.
The NLRI field in the UPDATE message is checked for syntactic validity. If the field is syntactically incorrect, then the Error Subcode is set to Invalid Network Field.
6.4 NOTIFICATION message error handling.
If a peer sends a NOTIFICATION message, and there is an error in that message, there is unfortunately no means of reporting this error via a subsequent NOTIFICATION message. Any such error, such as an unrecognized Error Code or Error Subcode, should be noticed, logged locally, and brought to the attention of the administration of the peer. The means to do this, however, lies outside the scope of this document.
6.5 Hold Timer Expired error handling.
If a system does not receive successive KEEPALIVE and/or UPDATE and/or NOTIFICATION messages within the period specified in the Hold Time field of the OPEN message, then the NOTIFICATION message with Hold Timer Expired Error Code must be sent and the BGP connection closed.
6.6 Finite State Machine error handling.
Any error detected by the BGP Finite State Machine (e.g., receipt of an unexpected event) is indicated by sending the NOTIFICATION message with Error Code Finite State Machine Error.
In absence of any fatal errors (that are indicated in this section), a BGP peer may choose at any given time to close its BGP connection by sending the NOTIFICATION message with Error Code Cease. However, the Cease NOTIFICATION message must not be used when a fatal error indicated by this section does exist.
6.8 Connection collision detection.
If a pair of BGP speakers try simultaneously to establish a TCP connection to each other, then two parallel connections between this pair of speakers might well be formed. We refer to this situation as connection collision. Clearly, one of these connections must be closed.
Based on the value of the BGP Identifier a convention is established for detecting which BGP connection is to be preserved when a collision does occur. The convention is to compare the BGP Identifiers of the peers involved in the collision and to retain only the connection initiated by the BGP speaker with the higher-valued BGP Identifier.
Upon receipt of an OPEN message, the local system must examine all of its connections that are in the OpenConfirm state. A BGP speaker may also examine connections in an OpenSent state if it knows the BGP Identifier of the peer by means outside of the protocol. If among these connections there is a connection to a remote BGP speaker whose BGP Identifier equals the one in the OPEN message, then the local system performs the following collision resolution procedure:
7. BGP Version Negotiation.
BGP speakers may negotiate the version of the protocol by making multiple attempts to open a BGP connection, starting with the highest version number each supports. If an open attempt fails with an Error Code OPEN Message Error, and an Error Subcode Unsupported Version Number, then the BGP speaker has available the version number it tried, the version number its peer tried, the version number passed by its peer in the NOTIFICATION message, and the version numbers that it supports. If the two peers do support one or more common versions, then this will allow them to rapidly determine the highest common version. In order to support BGP version negotiation, future versions of BGP must retain the format of the OPEN and NOTIFICATION messages.
8. BGP Finite State machine.
This section specifies BGP operation in terms of a Finite State Machine (FSM). Following is a brief summary and overview of BGP operations by state as determined by this FSM. A condensed version of the BGP FSM is found in Appendix 1.
9. UPDATE Message Handling
An UPDATE message may be received only in the Established state. When an UPDATE message is received, each field is checked for validity as specified in Section 6.3.
If an optional non-transitive attribute is unrecognized, it is quietly ignored. If an optional transitive attribute is unrecognized, the Partial bit (the third high-order bit) in the attribute flags octet is set to 1, and the attribute is retained for propagation to other BGP speakers.
If an optional attribute is recognized, and has a valid value, then, depending on the type of the optional attribute, it is processed locally, retained, and updated, if necessary, for possible propagation to other BGP speakers.
If the UPDATE message contains a non-empty WITHDRAWN ROUTES field, the previously advertised routes whose destinations (expressed as IP prefixes) contained in this field shall be removed from the Adj-RIB- In. This BGP speaker shall run its Decision Process since the previously advertised route is not longer available for use.
If the UPDATE message contains a feasible route, it shall be placed in the appropriate Adj-RIB-In, and the following additional actions shall be taken:
9.1 Decision Process
The Decision Process selects routes for subsequent advertisement by applying the policies in the local Policy Information Base (PIB) to the routes stored in its Adj-RIB-In. The output of the Decision Process is the set of routes that will be advertised to all peers; the selected routes will be stored in the local speaker's Adj-RIB- Out.
The selection process is formalized by defining a function that takes the attribute of a given route as an argument and returns a non- negative integer denoting the degree of preference for the route. The function that calculates the degree of preference for a given route shall not use as its inputs any of the following: the existence of other routes, the non-existence of other routes, or the path attributes of other routes. Route selection then consists of individual application of the degree of preference function to each feasible route, followed by the choice of the one with the highest degree of preference.
The Decision Process operates on routes contained in each Adj-RIB-In, and is responsible for:
The Decision Process takes place in three distinct phases, each triggered by a different event:
9.1.1 Phase 1: Calculation of Degree of Preference
The Phase 1 decision function shall be invoked whenever the local BGP speaker receives an UPDATE message from a peer located in a neighboring autonomous system that advertises a new route, a replacement route, or a withdrawn route.
The Phase 1 decision function is a separate process which completes when it has no further work to do.
The Phase 1 decision function shall lock an Adj-RIB-In prior to operating on any route contained within it, and shall unlock it after operating on all new or unfeasible routes contained within it.
For each newly received or replacement feasible route, the local BGP speaker shall determine a degree of preference. If the route is learned from a BGP speaker in the local autonomous system, either the value of the LOCAL_PREF attribute shall be taken as the degree of preference, or the local system shall compute the degree of preference of the route based on preconfigured policy information. If the route is learned from a BGP speaker in a neighboring autonomous system, then the degree of preference shall be computed based on preconfigured policy information. The exact nature of this policy information and the computation involved is a local matter. The local speaker shall then run the internal update process of 9.2.1 to select and advertise the most preferable route.
9.1.2 Phase 2: Route Selection
The Phase 2 decision function shall be invoked on completion of Phase 1. The Phase 2 function is a separate process which completes when it has no further work to do. The Phase 2 process shall consider all routes that are present in the Adj-RIBs-In, including those received from BGP speakers located in its own autonomous system and those received from BGP speakers located in neighboring autonomous systems.
The Phase 2 decision function shall be blocked from running while the Phase 3 decision function is in process. The Phase 2 function shall lock all Adj-RIBs-In prior to commencing its function, and shall unlock them on completion.
If the NEXT_HOP attribute of a BGP route depicts an address to which the local BGP speaker doesn't have a route in its Loc-RIB, the BGP route SHOULD be excluded from the Phase 2 decision function.
For each set of destinations for which a feasible route exists in the Adj-RIBs-In, the local BGP speaker shall identify the route that has:
The local speaker SHALL then install that route in the Loc-RIB, replacing any route to the same destination that is currently being held in the Loc-RIB. The local speaker MUST determine the immediate next hop to the address depicted by the NEXT_HOP attribute of the selected route by performing a lookup in the IGP and selecting one of the possible paths in the IGP. This immediate next hop MUST be used when installing the selected route in the Loc-RIB. If the route to the address depicted by the NEXT_HOP attribute changes such that the immediate next hop changes, route selection should be recalculated as specified above.
Unfeasible routes shall be removed from the Loc-RIB, and corresponding unfeasible routes shall then be removed from the Adj- RIBs-In.
188.8.131.52 Breaking Ties (Phase 2)
In its Adj-RIBs-In a BGP speaker may have several routes to the same destination that have the same degree of preference. The local speaker can select only one of these routes for inclusion in the associated Loc-RIB. The local speaker considers all equally preferable routes, both those received from BGP speakers located in neighboring autonomous systems, and those received from other BGP speakers located in the local speaker's autonomous system.
The following tie-breaking procedure assumes that for each candidate route all the BGP speakers within an autonomous system can ascertain the cost of a path (interior distance) to the address depicted by the NEXT_HOP attribute of the route. Ties shall be broken according to the following algorithm:
9.1.3 Phase 3: Route Dissemination
The Phase 3 decision function shall be invoked on completion of Phase 2, or when any of the following events occur:
The Phase 3 function is a separate process which completes when it has no further work to do. The Phase 3 Routing Decision function shall be blocked from running while the Phase 2 decision function is in process.
All routes in the Loc-RIB shall be processed into a corresponding entry in the associated Adj-RIBs-Out. Route aggregation and information reduction techniques (see 184.108.40.206) may optionally be applied.
For the benefit of future support of inter-AS multicast capabilities, a BGP speaker that participates in inter-AS multicast routing shall advertise a route it receives from one of its external peers and if it installs it in its Loc-RIB, it shall advertise it back to the peer from which the route was received. For a BGP speaker that does not participate in inter-AS multicast routing such an advertisement is optional. When doing such an advertisement, the NEXT_HOP attribute should be set to the address of the peer. An implementation may also optimize such an advertisement by truncating information in the AS_PATH attribute to include only its own AS number and that of the peer that advertised the route (such truncation requires the ORIGIN attribute to be set to INCOMPLETE). In addition an implementation is not required to pass optional or discretionary path attributes with such an advertisement.
When the updating of the Adj-RIBs-Out and the Forwarding Information Base (FIB) is complete, the local BGP speaker shall run the external update process of 9.2.2.
9.1.4 Overlapping Routes
A BGP speaker may transmit routes with overlapping Network Layer Reachability Information (NLRI) to another BGP speaker. NLRI overlap occurs when a set of destinations are identified in non-matching multiple routes. Since BGP encodes NLRI using IP prefixes, overlap will always exhibit subset relationships. A route describing a smaller set of destinations (a longer prefix) is said to be more specific than a route describing a larger set of destinations (a shorted prefix); similarly, a route describing a larger set of destinations (a shorter prefix) is said to be less specific than a route describing a smaller set of destinations (a longer prefix).
The precedence relationship effectively decomposes less specific routes into two parts:
When overlapping routes are present in the same Adj-RIB-In, the more specific route shall take precedence, in order from more specific to least specific.
The set of destinations described by the overlap represents a portion of the less specific route that is feasible, but is not currently in use. If a more specific route is later withdrawn, the set of destinations described by the overlap will still be reachable using the less specific route.
If a BGP speaker receives overlapping routes, the Decision Process shall take into account the semantics of the overlapping routes. In particular, if a BGP speaker accepts the less specific route while rejecting the more specific route from the same peer, then the destinations represented by the overlap may not forward along the ASs listed in the AS_PATH attribute of that route. Therefore, a BGP speaker has the following choices:
If a BGP speaker chooses e), then it should add ATOMIC_AGGREGATE attribute to the route. A route that carries ATOMIC_AGGREGATE attribute can not be de-aggregated. That is, the NLRI of this route can not be made more specific. Forwarding along such a route does not guarantee that IP packets will actually traverse only ASs listed in the AS_PATH attribute of the route. If a BGP speaker chooses a), it must not advertise the more general route without the more specific route.
9.2 Update-Send Process
The Update-Send process is responsible for advertising UPDATE messages to all peers. For example, it distributes the routes chosen by the Decision Process to other BGP speakers which may be located in either the same autonomous system or a neighboring autonomous system. rules for information exchange between BGP speakers located in different autonomous systems are given in 9.2.2; rules for information exchange between BGP speakers located in the same autonomous system are given in 9.2.1.
Distribution of routing information between a set of BGP speakers, all of which are located in the same autonomous system, is referred to as internal distribution.
9.2.1 Internal Updates
The Internal update process is concerned with the distribution of routing information to BGP speakers located in the local speaker's autonomous system.
When a BGP speaker receives an UPDATE message from another BGP speaker located in its own autonomous system, the receiving BGP speaker shall not re-distribute the routing information contained in that UPDATE message to other BGP speakers located in its own autonomous system.
When a BGP speaker receives a new route from a BGP speaker in a neighboring autonomous system, it shall advertise that route to all other BGP speakers in its autonomous system by means of an UPDATE message if any of the following conditions occur:
When a BGP speaker receives an UPDATE message with a non-empty WITHDRAWN ROUTES field, it shall remove from its Adj-RIB-In all routes whose destinations was carried in this field (as IP prefixes). The speaker shall take the following additional steps:
All feasible routes which are advertised shall be placed in the appropriate Adj-RIBs-Out, and all unfeasible routes which are advertised shall be removed from the Adj-RIBs-Out.
220.127.116.11 Breaking Ties (Internal Updates)
If a local BGP speaker has connections to several BGP speakers in neighboring autonomous systems, there will be multiple Adj-RIBs-In associated with these peers. These Adj-RIBs-In might contain several equally preferable routes to the same destination, all of which were advertised by BGP speakers located in neighboring autonomous systems. The local BGP speaker shall select one of these routes according to the following rules:
9.2.2 External Updates
The external update process is concerned with the distribution of routing information to BGP speakers located in neighboring autonomous systems. As part of Phase 3 route selection process, the BGP speaker has updated its Adj-RIBs-Out and its Forwarding Table. All newly installed routes and all newly unfeasible routes for which there is no replacement route shall be advertised to BGP speakers located in neighboring autonomous systems by means of UPDATE message.
Any routes in the Loc-RIB marked as unfeasible shall be removed. Changes to the reachable destinations within its own autonomous system shall also be advertised in an UPDATE message.
9.2.3 Controlling Routing Traffic Overhead
The BGP protocol constrains the amount of routing traffic (that is, UPDATE messages) in order to limit both the link bandwidth needed to advertise UPDATE messages and the processing power needed by the Decision Process to digest the information contained in the UPDATE messages.
18.104.22.168 Frequency of Route Advertisement
The parameter MinRouteAdvertisementInterval determines the minimum amount of time that must elapse between advertisement of routes to a particular destination from a single BGP speaker. This rate limiting procedure applies on a per-destination basis, although the value of MinRouteAdvertisementInterval is set on a per BGP peer basis.
Two UPDATE messages sent from a single BGP speaker that advertise feasible routes to some common set of destinations received from BGP speakers in neighboring autonomous systems must be separated by at least MinRouteAdvertisementInterval. Clearly, this can only be achieved precisely by keeping a separate timer for each common set of destinations. This would be unwarranted overhead. Any technique which ensures that the interval between two UPDATE messages sent from a single BGP speaker that advertise feasible routes to some common set of destinations received from BGP speakers in neighboring autonomous systems will be at least MinRouteAdvertisementInterval, and will also ensure a constant upper bound on the interval is acceptable.
Since fast convergence is needed within an autonomous system, this procedure does not apply for routes receives from other BGP speakers in the same autonomous system. To avoid long-lived black holes, the procedure does not apply to the explicit withdrawal of unfeasible routes (that is, routes whose destinations (expressed as IP prefixes) are listed in the WITHDRAWN ROUTES field of an UPDATE message).
This procedure does not limit the rate of route selection, but only the rate of route advertisement. If new routes are selected multiple times while awaiting the expiration of MinRouteAdvertisementInterval, the last route selected shall be advertised at the end of MinRouteAdvertisementInterval.
22.214.171.124 Frequency of Route Origination
The parameter MinASOriginationInterval determines the minimum amount of time that must elapse between successive advertisements of UPDATE messages that report changes within the advertising BGP speaker's own autonomous systems.
To minimize the likelihood that the distribution of BGP messages by a given BGP speaker will contain peaks, jitter should be applied to the timers associated with MinASOriginationInterval, Keepalive, and MinRouteAdvertisementInterval. A given BGP speaker shall apply the same jitter to each of these quantities regardless of the destinations to which the updates are being sent; that is, jitter will not be applied on a "per peer" basis.
The amount of jitter to be introduced shall be determined by multiplying the base value of the appropriate timer by a random factor which is uniformly distributed in the range from 0.75 to 1.0.
9.2.4 Efficient Organization of Routing Information
Having selected the routing information which it will advertise, a BGP speaker may avail itself of several methods to organize this information in an efficient manner.
126.96.36.199 Information Reduction
Information reduction may imply a reduction in granularity of policy control - after information is collapsed, the same policies will apply to all destinations and paths in the equivalence class.
The Decision Process may optionally reduce the amount of information that it will place in the Adj-RIBs-Out by any of the following methods:
Destination IP addresses can be represented as IP address prefixes. In cases where there is a correspondence between the address structure and the systems under control of an autonomous system administrator, it will be possible to reduce the size of the NLRI carried in the UPDATE messages.
AS path information can be represented as ordered AS_SEQUENCEs or unordered AS_SETs. AS_SETs are used in the route aggregation algorithm described in 188.8.131.52. They reduce the size of the AS_PATH information by listing each AS number only once, regardless of how many times it may have appeared in multiple AS_PATHs that were aggregated.
An AS_SET implies that the destinations listed in the NLRI can be reached through paths that traverse at least some of the constituent autonomous systems. AS_SETs provide sufficient information to avoid routing information looping; however their use may prune potentially feasible paths, since such paths are no longer listed individually as in the form of AS_SEQUENCEs. In practice this is not likely to be a problem, since once an IP packet arrives at the edge of a group of autonomous systems, the BGP speaker at that point is likely to have more detailed path information and can distinguish individual paths to destinations.
184.108.40.206 Aggregating Routing Information
Aggregation is the process of combining the characteristics of several different routes in such a way that a single route can be advertised. Aggregation can occur as part of the decision process to reduce the amount of routing information that will be placed in the Adj-RIBs-Out.
Aggregation reduces the amount of information that a BGP speaker must store and exchange with other BGP speakers. Routes can be aggregated by applying the following procedure separately to path attributes of like type and to the Network Layer Reachability Information.
Routes that have the following attributes shall not be aggregated unless the corresponding attributes of each route are identical: MULTI_EXIT_DISC, NEXT_HOP.
Path attributes that have different type codes can not be aggregated together. Path of the same type code may be aggregated, according to the following rules:
ORIGIN attribute: If at least one route among routes that are aggregated has ORIGIN with the value INCOMPLETE, then the aggregated route must have the ORIGIN attribute with the value INCOMPLETE. Otherwise, if at least one route among routes that are aggregated has ORIGIN with the value EGP, then the aggregated route must have the origin attribute with the value EGP. In all other case the value of the ORIGIN attribute of the aggregated route is INTERNAL.
AS_PATH attribute: If routes to be aggregated have identical AS_PATH attributes, then the aggregated route has the same AS_PATH attribute as each individual route.
For the purpose of aggregating AS_PATH attributes we model each AS within the AS_PATH attribute as a tuple <type, value>, where "type" identifies a type of the path segment the AS belongs to
(e.g. AS_SEQUENCE, AS_SET), and "value" is the AS number. If the routes to be aggregated have different AS_PATH attributes, then the aggregated AS_PATH attribute shall satisfy all of the following conditions:
An implementation may choose any algorithm which conforms to these rules. At a minimum a conformant implementation shall be able to perform the following algorithm that meets all of the above conditions:
Appendix 6, section 6.8 presents another algorithm that satisfies the conditions and allows for more complex policy configurations.
ATOMIC_AGGREGATE: If at least one of the routes to be aggregated has ATOMIC_AGGREGATE path attribute, then the aggregated route shall have this attribute as well.
AGGREGATOR: All AGGREGATOR attributes of all routes to be aggregated should be ignored.
9.3 Route Selection Criteria
Generally speaking, additional rules for comparing routes among several alternatives are outside the scope of this document. There are two exceptions:
9.4 Originating BGP routes
A BGP speaker may originate BGP routes by injecting routing information acquired by some other means (e.g. via an IGP) into BGP. A BGP speaker that originates BGP routes shall assign the degree of preference to these routes by passing them through the Decision Process (see Section 9.1). These routes may also be distributed to other BGP speakers within the local AS as part of the Internal update process (see Section 9.2.1). The decision whether to distribute non- BGP acquired routes within an AS via BGP or not depends on the environment within the AS (e.g. type of IGP) and should be controlled via configuration.
Appendix 1. BGP FSM State Transitions and Actions.
This Appendix discusses the transitions between states in the BGP FSM in response to BGP events. The following is the list of these states and events when the negotiated Hold Time value is non-zero.
The following table describes the state transitions of the BGP FSM and the actions triggered by these transitions.
Message Sent Next
The following is a condensed version of the above state transition table.
Appendix 2. Comparison with RFC1267
BGP-4 is capable of operating in an environment where a set of reachable destinations may be expressed via a single IP prefix. The concept of network classes, or subnetting is foreign to BGP-4. To accommodate these capabilities BGP-4 changes semantics and encoding associated with the AS_PATH attribute. New text has been added to define semantics associated with IP prefixes. These abilities allow BGP-4 to support the proposed supernetting scheme .
To simplify configuration this version introduces a new attribute, LOCAL_PREF, that facilitates route selection procedures.
The INTER_AS_METRIC attribute has been renamed to be MULTI_EXIT_DISC. A new attribute, ATOMIC_AGGREGATE, has been introduced to insure that certain aggregates are not de-aggregated. Another new attribute, AGGREGATOR, can be added to aggregate routes in order to advertise which AS and which BGP speaker within that AS caused the aggregation.
To insure that Hold Timers are symmetric, the Hold Time is now negotiated on a per-connection basis. Hold Times of zero are now supported.
Appendix 3. Comparison with RFC 1163
All of the changes listed in Appendix 2, plus the following.
To detect and recover from BGP connection collision, a new field (BGP Identifier) has been added to the OPEN message. New text (Section 6.8) has been added to specify the procedure for detecting and recovering from collision.
The new document no longer restricts the border router that is passed in the NEXT_HOP path attribute to be part of the same Autonomous System as the BGP Speaker.
New document optimizes and simplifies the exchange of the information about previously reachable routes.
Appendix 4. Comparison with RFC 1105
All of the changes listed in Appendices 2 and 3, plus the following.
Minor changes to the RFC1105 Finite State Machine were necessary to accommodate the TCP user interface provided by 4.3 BSD.
The notion of Up/Down/Horizontal relations present in RFC1105 has been removed from the protocol.
The changes in the message format from RFC1105 are as follows:
Note that quite often BGP, as specified in RFC 1105, is referred to as BGP-1, BGP, as specified in RFC 1163, is referred to as BGP-2, BGP, as specified in RFC1267 is referred to as BGP-3, and BGP, as specified in this document is referred to as BGP-4.
Appendix 5. TCP options that may be used with BGP
If a local system TCP user interface supports TCP PUSH function, then each BGP message should be transmitted with PUSH flag set. Setting PUSH flag forces BGP messages to be transmitted promptly to the receiver.
If a local system TCP user interface supports setting precedence for TCP connection, then the BGP transport connection should be opened with precedence set to Internetwork Control (110) value (see also ).
Appendix 6. Implementation Recommendations
This section presents some implementation recommendations.
6.1 Multiple Networks Per Message
The BGP protocol allows for multiple address prefixes with the same AS path and next-hop gateway to be specified in one message. Making use of this capability is highly recommended. With one address prefix per message there is a substantial increase in overhead in the receiver. Not only does the system overhead increase due to the reception of multiple messages, but the overhead of scanning the routing table for updates to BGP peers and other routing protocols (and sending the associated messages) is incurred multiple times as well. One method of building messages containing many address prefixes per AS path and gateway from a routing table that is not organized per AS path is to build many messages as the routing table is scanned. As each address prefix is processed, a message for the associated AS path and gateway is allocated, if it does not exist, and the new address prefix is added to it. If such a message exists, the new address prefix is just appended to it. If the message lacks the space to hold the new address prefix, it is transmitted, a new message is allocated, and the new address prefix is inserted into the new message. When the entire routing table has been scanned, all allocated messages are sent and their resources released. Maximum compression is achieved when all the destinations covered by the address prefixes share a gateway and common path attributes, making it possible to send many address prefixes in one 4096-byte message.
When peering with a BGP implementation that does not compress multiple address prefixes into one message, it may be necessary to take steps to reduce the overhead from the flood of data received when a peer is acquired or a significant network topology change occurs. One method of doing this is to limit the rate of updates. This will eliminate the redundant scanning of the routing table to provide flash updates for BGP peers and other routing protocols. A disadvantage of this approach is that it increases the propagation latency of routing information. By choosing a minimum flash update interval that is not much greater than the time it takes to process the multiple messages this latency should be minimized. A better method would be to read all received messages before sending updates.
6.2 Processing Messages on a Stream Protocol
BGP uses TCP as a transport mechanism. Due to the stream nature of TCP, all the data for received messages does not necessarily arrive at the same time. This can make it difficult to process the data as messages, especially on systems such as BSD Unix where it is not possible to determine how much data has been received but not yet processed.
One method that can be used in this situation is to first try to read just the message header. For the KEEPALIVE message type, this is a complete message; for other message types, the header should first be verified, in particular the total length. If all checks are successful, the specified length, minus the size of the message header is the amount of data left to read. An implementation that would "hang" the routing information process while trying to read from a peer could set up a message buffer (4096 bytes) per peer and fill it with data as available until a complete message has been received.
6.3 Reducing route flapping
To avoid excessive route flapping a BGP speaker which needs to withdraw a destination and send an update about a more specific or less specific route shall combine them into the same UPDATE message.
6.4 BGP Timers
BGP employs five timers: ConnectRetry, Hold Time, KeepAlive, MinASOriginationInterval, and MinRouteAdvertisementInterval The suggested value for the ConnectRetry timer is 120 seconds. The suggested value for the Hold Time is 90 seconds. The suggested value for the KeepAlive timer is 30 seconds. The suggested value for the MinASOriginationInterval is 15 seconds. The suggested value for the MinRouteAdvertisementInterval is 30 seconds.
An implementation of BGP MUST allow these timers to be configurable.
6.5 Path attribute ordering
Implementations which combine update messages as described above in 6.1 may prefer to see all path attributes presented in a known order. This permits them to quickly identify sets of attributes from different update messages which are semantically identical. To facilitate this, it is a useful optimization to order the path attributes according to type code. This optimization is entirely optional.
6.6 AS_SET sorting
Another useful optimization that can be done to simplify this situation is to sort the AS numbers found in an AS_SET. This optimization is entirely optional.
6.7 Control over version negotiation
Since BGP-4 is capable of carrying aggregated routes which cannot be properly represented in BGP-3, an implementation which supports BGP-4 and another BGP version should provide the capability to only speak BGP-4 on a per-peer basis.
6.8 Complex AS_PATH aggregation
An implementation which chooses to provide a path aggregation algorithm which retains significant amounts of path information may wish to use the following procedure:
If as a result of the above procedure a given AS number appears more than once within the aggregated AS_PATH attribute, all, but the last instance (rightmost occurrence) of that AS number should be removed from the aggregated AS_PATH attribute.
 Mills, D., "Exterior Gateway Protocol Formal Specification", RFC 904, BBN, April 1984.
 Rekhter, Y., "EGP and Policy Based Routing in the New NSFNET Backbone", RFC 1092, T.J. Watson Research Center, February 1989.
 Braun, H-W., "The NSFNET Routing Architecture", RFC 1093, MERIT/NSFNET Project, February 1989.
 Postel, J., "Transmission Control Protocol - DARPA Internet Program Protocol Specification", STD 7, RFC 793, DARPA, September 1981.
 Rekhter, Y., and P. Gross, "Application of the Border Gateway Protocol in the Internet", RFC 1772, T.J. Watson Research Center, IBM Corp., MCI, March 1995.
 Postel, J., "Internet Protocol - DARPA Internet Program Protocol Specification", STD 5, RFC 791, DARPA, September 1981.
 "Information Processing Systems - Telecommunications and Information Exchange between Systems - Protocol for Exchange of Inter-domain Routeing Information among Intermediate Systems to Support Forwarding of ISO 8473 PDUs", ISO/IEC IS10747, 1993
 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
 Rekhter, Y., Li, T., "An Architecture for IP Address Allocation with CIDR", RFC 1518, T.J. Watson Research Center, cisco, September 1993
Security issues are not discussed in this document.
Yakov Rekhter T.J. Watson Research Center IBM Corporation P.O. Box 704, Office H3-D40 Yorktown Heights, NY 10598
Phone: +1 914 784 7361 EMail: email@example.com
Tony Li cisco Systems, Inc. 170 W. Tasman Dr. San Jose, CA 95134
Rekhter & Li [Page 57]