Single-Homed EBGP
927
Characteristic
BGP
OSPF
Metric
Various
Bandwidth
Hierarchy
No
Yes
Building blocks
Autonomous systems
Areas
Base protocol
TCP port 179
Protocol value 89
Traffic type
Unicast
Multicast
Neighbors
Specifically configured
Discovered/configured
Route exchange
Only with neighbors
Only with adjacent neighbors
Initial update
Synchronize database
Synchronize database
Update frequency
Incremental
Incremental with 60-minute timer
Hello timer
60 seconds
10 or 30 seconds
Hold timer
180 seconds
40 or 120 seconds
Internal route exchange
Internal BGP sessions
LSA Types 1 and 2
External route exchange
External BGP sessions
LSA Types 3, 4, and 5
Route updates
Contain network, attributes,
AS path
Contain network, metric (Types
3 and 4 LSAs)
Network statement
Advertises network
Activates OSPF on interface
Special features
Route reflectors
Stub, totally stubby, NSSA areas
It’s easy to get lost in the specific details of BGP, so at the risk of a small amount of
repetition later on, the following is a high-level overview of the BGP protocol and its main
characteristics as listed in Table 21.1.
BGP is a distance-vector protocol, which means that it advertises all or a portion of its
route table to its neighbors. The advertised routes include the network being advertised, a
list of attributes that influence the selection of the best path, the next-hop address through
which the network can be reached, and a list of autonomous systems (ASs) through which
the route update has passed. BGP routers use the list of autonomous systems to ensure a
928
Chapter 21
■
Wide Area Networks
loop-free path by enforcing the rule that no AS path list is allowed to contain the same AS
number twice.
BGP supports classless networks, the use of variable length subnet masks (VLSMs),
and summarization. These characteristics allow BGP to work with networks that are not
organized on purely classful boundaries and to create summaries of networks to reduce the
size of the routing tables.
■
BGP uses a rich variety of metrics called attributes to influence the selection of the
best path to remote networks in the event that there are multiple advertised paths.
Network administrators far removed from the initial origin of the advertised networks
can manipulate these attributes. Paths can be chosen simply because a neighbor AS is
preferred for political or economic reasons, thus overriding more traditional measures
such as the distance to the advertised route.
■
BGP supports a nonhierarchical network structure and allows a complex combination
of interconnections among neighbors. There is no counterpart in BGP to the OSPF
concept of area 0, which is the single area through which interarea traffic passes. Traffic
between different BGP autonomous systems may follow a variety of different paths.
■
BGP uses the concept of autonomous systems to define the boundaries of networks
and treats communications among neighbors differently depending on whether the
neighbors belong to the same autonomous system or not. An autonomous system is a
collection of routers that are under a common administrative control and that present a
common route policy to the outside world. It is not necessary that all of the routers run
the same routing protocol, just that they all be controlled and coordinated by the same
administrative authority.
■
An AS uses BGP to advertise routes that are in its network and need to be visible
outside of the network; it also uses BGP to learn about the reachability of routes by
listening to advertisement announcements from other autonomous systems. Each AS
can have a specific policy regarding the routes it wishes to advertise externally. These
policies can be different for every point in which the AS attaches to the outside world.
■
Inside autonomous networks, interior gateway protocols (IGPs) are used to discover
the connectivity among a set of IP subnets. IGPs are well-known protocols such as the
Routing Information Protocol (RIP), Interior Gateway Routing Protocol (IGRP), Open
Shortest Path First (OSPF), and Enhanced Interior Gateway Routing Protocol (EIGRP).
BGP relies upon TCP for connection-oriented, acknowledged communications using
port 179. BGP routers are specifically configured as neighbors of one another and use
unicast packets to exchange route information, keepalives, and a variety of other messages.
BGP routers go through a variety of stages as they establish communications with their
configured neighbors, verify consistent parameter configuration, and begin the initial syn-
chronization of their route information. After the initial synchronization is complete, BGP
neighbors exchange updates on a triggered basis and monitor their connection state via
periodic keepalives.
BGP neighbors either live in the same AS, in which case they are referred to as internal BGP
(iBGP) neighbors, or live in different ASs, in which case they are referred to as
external BGP
Single-Homed EBGP
929
(eBGP) neighbors. Internal BGP neighbors do not need to share a common network and can
be separated by many other routers that don’t need to run BGP. However, every iBGP router
must be configured as a neighbor to every other iBGP router in the same area. External BGP
neighbors are normally required to share a common network and are directly accessible to each
other. There is no requirement that every eBGP router be a neighbor to every other eBGP router.
BGP can advertise networks that are learned dynamically, statically, or through redistribution.
The network command, which is used in most other protocols to cause the router’s interface(s) to
begin listening for and sending route updates, is used in BGP routers to advertise specific networks.
There are a number of rules, like the synchronization rule, that govern BGP’s interaction with inte-
rior gateway protocols and the routes that are advertised as a result of this interaction.
Finally, BGP can implement a variety of mechanisms to improve scalability.
Summarization is certainly one of these mechanisms, as is the use of route reflectors. Route
reflectors provide a means to eliminate the requirement for a full mesh of neighbor relation-
ships among iBGP neighbors, thus permitting larger BGP environments with less traffic.
Do'stlaringiz bilan baham: 
