Cisco Systems 3560 Manual De Usuario

detailed information about BGP in Internet Routing Architectures, published by Cisco Press, and in the 
“Configuring BGP” chapter in the Cisco IP and IP Routing Configuration Guide from the Cisco.com 
page under Documentation > Cisco IOS Software > 12.2 Mainline > Configuration Guides. 

For details about BGP commands and keywords, see the “IP Routing Protocols” part of the Cisco IOS 
IP Command Reference, Volume 2 of 3: Routing Protocols, Release 12.2 under Documentation > Cisco 
IOS Software > 12.2 Mainline > Command References. For a list of BGP commands that are visible 
but not supported by the switch, see 

Routers that belong to the same autonomous system (AS) and that exchange BGP updates run internal 
BGP (IBGP), and routers that belong to different autonomous systems and that exchange BGP updates 
run external BGP (EBGP). Most configuration commands are the same for configuring EBGP and IBGP. 
The difference is that the routing updates are exchanged either between autonomous systems (EBGP) or 
within an AS (IBGP). 

 shows a network that is running both EBGP and IBGP.

Before exchanging information with an external AS, BGP ensures that networks within the AS can be 
reached by defining internal BGP peering among routers within the AS and by redistributing BGP 
routing information to IGPs that run within the AS, such as IGRP and OSPF.

Routers that run a BGP routing process are often referred to as BGP speakers. BGP uses the 
Transmission Control Protocol (TCP) as its transport protocol (specifically port 179). Two BGP speakers 
that have a TCP connection to each other for exchanging routing information are known as peers or 
neighbors. In 

, Routers A and B are BGP peers, as are Routers B and C and Routers C and 

D. The routing information is a series of AS numbers that describe the full path to the destination 
network. BGP uses this information to construct a loop-free map of autonomous systems.

The network has these characteristics:

Routers A and B are running EBGP, and Routers B and C are running IBGP. Note that the EBGP 
peers are directly connected and that the IBGP peers are not. As long as there is an IGP running that 
allows the two neighbors to reach one another, IBGP peers do not have to be directly connected. 

All BGP speakers within an AS must establish a peer relationship with each other. That is, the BGP 
speakers within an AS must be fully meshed logically. BGP4 provides two techniques that reduce 
the requirement for a logical full mesh: confederations and route reflectors.

AS 200 is a transit AS for AS 100 and AS 300—that is, AS 200 is used to transfer packets between 
AS 100 and AS 300.

AS 100

74775

AS 200

129.213.1.2

175.220.212.1

129.213.1.1

192.208.10.2

AS 300

EBGP

EBGP

192.208.10.1

175.220.1.2

IBGP

Router B

Router A

Router D

Router C