Bgp - Allied Telesis SwitchBlade x908 Series Software Reference Manual

During boot-up, an OSPF router initializes its routing-protocol-specific data structures and
tables. When the lower-layer protocols with which it interfaces are functional, it sends the
OSPF Hello protocol packets to find neighboring routers. A router sends Hello packets as
keep-alive packets, informing other routers about its continuing functionality. Two routers are
adjacent when their link state databases are synchronized.
Multi-access networks have more than two routers. On multi-access networks, the hello
protocol chooses a designated router and a designated backup-router. The designated router
generates LSAs for the entire multi-access network, and reduces network traffic and the size of
the topological database. The designated router also determines the adjacency of routers and
the synchronization of their topological databases. The data on a router's adjacencies or state
changes are provided by periodic transmission of an LSA. Failed routers are detected, and
topology is changed quickly by comparison of adjacencies to link states. Each router calculates a
shortest path tree, with itself as a root, from the topological database generated from these
LSAs. This shortest path tree creates a routing table.

BGP

Border Gateway Protocol (BGP) is an exterior gateway protocol (EGP) that determines the
best path in networks, performs optimal routing between multiple autonomous systems or
domains, and exchanges routing information with other BGP systems. The RFCs 1771 (BGP4),
1654 (first BGP4 specification), and 1105, 1163, 1267 (older version of BGP) describe BGP and
BGP4.
Multiple-peer BGP routers in different autonomous systems, or administrative domains on the
same physical network, support consistent internetwork topology using inter-autonomous
system routing. Multiple-peer BGP routers within the same AS support consistent system
topology using inter-autonomous system routing. BGP determines the router to serve as the
connection point for specific external autonomous system routing services. Multiple-peer BGP
routers transport traffic across an autonomous system that does not run BGP using pass-
through autonomous-systems routing. In this case, if the traffic does not originate, or is destined
for an autonomous system under consideration, the AS is used only to transport (pass-through)
the traffic using another intra-autonomous system routing protocol.
BGP exchanges information about the list of autonomous system paths with other BGP
systems. A connectivity mapping between autonomous systems is created, routing loops are
pruned, and other autonomous systems-level policy decisions are taken. Each BGP router
maintains a routing table of all feasible and optimal paths to other networks and incrementally
updates the routing information received from other peer BGP routers.
The BGP routing metric describes the preference of the path, and is assigned to each link by
the network administrator. The network administrator assigns this value to a link depending on
path criteria such as:
■
■
■
■
■
C613-50007-01 REV B
the number of autonomous systems through which the path passes
the history of stability
the line speed
any delays
cost per packet
Software Reference for SwitchBlade® x908, x900 and x600 Series Switches
TM
AlliedWare Plus Operating System - Software Version 5.3.1
Routing Protocol Overview
28.3