User Manual
UMN:CLI
V8102
595
12 IP Routing Protocol
12.1
Border Gateway Protocol (BGP)
The Border Gateway Protocol (BGP) is an exterior gateway protocol (EGP) that is used to
exchange routing information among routers in different autonomous systems (AS). BGP
routing information includes the complete route to each destination. BGP uses the routing
information to maintain a database of network reachability information, which it exchang-
es with other BGP systems. BGP uses the network reachability information to construct a
graph of AS connectivity, thus allowing BGP to remove routing loops and en-force policy
decisions at the AS level.
Multiprotocol BGP (MBGP) extensions enable BGP to support IPv6. MBGP defines the
attributes MP_REACH_NLRI and MP_UNREACH_NLRI, which are used to carry IPv6
reachability information. Network layer reachability information (NLRI) update messages
carry IPv6 address prefixes of feasible routes.
BGP allows for policy-based routing. You can use routing policies to choose among multi-
ple paths to a destination and to control the redistribution of routing information.
BGP uses the Transmission Control Protocol (TCP) as its transport protocol, using port
179 for establishing connections. Running over a reliable transport protocol eliminates the
need for BGP to implement update fragmentation, retransmission, acknowledgment, and
sequencing.
The routing protocol software supports BGP version 4. This version of BGP adds support
for classless interdomain routing (CIDR), which eliminates the concept of network classes.
Instead of assuming which bits of an address represent the network by looking at the first
octet, CIDR allows you to explicitly specify the number of bits in the network address,
thus providing a means to decrease the size of the routing tables. BGP version 4 also
supports aggregation of routes, including the aggregation of AS paths
An Autonomous System (AS) is a set of routers that are under a single technical admin-
istration and normally use a single interior gateway protocol and a common set of metrics
to propagate routing information within the set of routers. To other ASs, an AS appears to
have a single, coherent interior routing plan and presents a consistent picture of what
destinations are reachable through it.
The two most important consequences are the need for interior routing protocols to reach
one hop beyond the AS boundary, and for BGP sessions to be fully meshed within an AS.
Since the next-hop contains the IP address of a router interface in the next autonomous
system, and this IP address is used to perform routing, the interior routing protocol must
be able to route to this address. This means that interior routing tables must include en-
tries one hop beyond the AS boundary. When a BGP routing update is received from a
neighboring AS, it must be relayed directly to all other BGP speakers in the AS. Do not
expect to relay BGP paths from one router, through another, to a third, all within the same
AS.