Multiple autonomous systems

FIGURE 8 . 1
Multiple autonomous systems
AS
96
AS 7
AS
28
AS
100
AS
20000
AS
57
AS
10
AS
32962
AS
29952
AS 1
AS 2 AS 3
BGP Operation
239
Once BGP speakers have formed peers and created their BGP tree, they will start exchanging
routing information. The BGP speakers will first exchange their entire BGP routing tables. From
that point forward, the peers will exchange incremental updates of their BGP routing tables and
KEEPALIVE messages to keep the connection up.
Understand that the BGP routing table is actually a new structure. It is not the
IP routing table, now that BGP entries are found in it. Instead, the BGP routing
table is much like a topology database, and it contains entries that may never
make it into the IP routing table for one reason or another. While the command
show ip route bgp
will display any BGP-learned routes that make it into the IP
routing table, the command
show ip bgp
is required to display the contents of
the actual BGP routing table.
From a bird’s-eye view, that’s all there is to BGP. Sounds a little too easy, right? That would be
a good assumption. BGP is a very complex routing protocol. That’s why we have dedicated two
chapters to it. We will now dive into the different components that make up the way BGP operates.