3Com Switch 7754, Configuration Manual

Page: 302 / 940

302 C
HAPTER
34: OSPF C
ONFIGURATION
topology of the whole network. Obviously, all routers get exactly the same
map.
■
A router uses the shortest path first (SPF) algorithm to calculate the shortest
path tree with itself as the root. The tree shows the routes to the nodes in the
autonomous system. External routes are leaf nodes, which are marked with the
routers from which they are advertised to record information outside the AS.
Obviously, the routing tables obtained by different routers are different.
Furthermore, to enable individual routers to broadcast their local status
information (such as available interface information and reachable neighbor
information) to the whole AS, routers in the AS should establish neighboring
relationship among them. In this case, the route changes on any router will result
in multiple transmissions, which are unnecessary and waste the precious
bandwidth resources. To solve this problem, designated router (DR) and backup
designated router (BDR) are defined in OSPF. For details about DR and BDR, see
“DR and BDR” on page 304.
OSPF supports interface-based packet authentication to guarantee the security of
route calculation. In addition, it transmits and receives packets in multicast
(224.0.0.5 and 224.0.0.6).
Basic OSPF Concepts Router ID
To run OSPF, a router must have a router ID. A router ID can be configured
manually. If no router ID is configured, the system will automatically select an IP
address from the IP addresses of the interfaces as the router ID. A router ID is
selected in the following way: if loopback interface addresses are configured, the
system chooses the latest configured IP address as the router ID; if no loopback
interface is configured, the first configured IP address among the IP addresses of
other interfaces will be the router ID.
DR and BDR
For details, see “DR and BDR” on page 304.
Area
If all the routers on an ever-growing huge network run OSPF, the large number of
routers will result in an enormous LSDB, which will consume an enormous storage
space, complicate the running of SPF algorithm, and increase CPU load.
Furthermore, as a network grows larger, it is more potential to have changes in the
network topology. Hence, the network will often be in “turbulence”, and a great
number of OSPF packets will be generated and transmitted in the network. This
will lower the network bandwidth utilization. In addition, each change will cause
all the routers on the network re-perform route calculation.
OSPF solves the above-mentioned problem by dividing an AS into multiple areas.
Areas group routers logically. A router on the border of an area belongs to more
than one area. A router connecting the backbone area to a non-backbone area is
called an area border router (ABR). An ABR can connect to the backbone area
physically or logically.
Area partition in OSPF reduces the number of LSAs in the network and enhances
OSPF scalability. To further reduce routing table size and the number of LSAs in