A routing table need not contain all routes in the network to reach all destinations. This simplification is arrived at through the use of a default route . When a router does not have an explicit route to a destination IP address, it looks to see if it has a default route in its routing table and, if so, forwards packets for this destination via the default route.

In RIP, the default route is represented as the IP address 0.0.0.0. This is convenient because 0.0.0.0 cannot be confused with any Class A, B, or C IP address.

One situation in which default routes can be employed in an intranet is in a core network that has branch offices hanging off it (Figure 2-7).

Figure 2-7. Branch offices only need a default route

Consider the topology of this figure. Since the branch offices have only one connection (to the core), all routes to the core network and to other branches can be replaced with a single default route pointing toward the core network. This implies that the size of the routing table in the branch offices is just the number of directly connected networks plus the default route.

So, router Portland may be configured as follows:

hostname Portland
...
interface Ethernet 0
ip address 192.100.1.17 255.255.255.240
!
interface Serial 0
ip address 192.100.1.33 255.255.255.240
!
router rip
network 192.100.1.0

An examination of Portland’s routing table would show the following:

Portland>sh ip route
...
Gateway of last resort is not set

     192.100.1.0/28 is subnetted, 2 subnets
C       192.100.1.16 is directly connected, Ethernet0
C       192.100.1.32 is directly connected, Serial0
R    0.0.0.0 [120/1] via 192.199.1.34, 0:00:21, Serial0

The default route may be sourced from router core1 as follows:

hostname core1
...
interface Serial 0
ip address 192.100.1.34 255.255.255.240
!
router rip
network 192.100.1.0
!
ip route 0.0.0.0 0.0.0.0 null0

Note that the default route 0.0.0.0 is automatically carried by RIP -- it is not listed in a network number statement under router rip.

The advantage of using a default in place of hundreds or thousands of more specific routes is obvious -- network bandwidth and router CPU are not tied up in routing updates. The disadvantage of using a default is that packets for destinations that are down or not even defined in the network are still forwarded to the core network.

Default routes are tremendously useful in Internet connectivity -- where all (thousands and thousands of ) Internet routes may be represented by a single default route.

Yet another use of default routes is in maintaining reachability between a routing domain running RIP and another routing domain with VLSM. Since VLSM cannot be imported into RIP, a default route pointing to the second domain may be defined in the RIP network.

Routes to hosts

Some host machines listen to RIP updates in “quiet” or “silent” mode (Figure 2-8). These hosts do not respond to requests for RIP routes or issue regular RIP updates. Listening to RIP provides redundancy to the hosts in a scenario in which multiple routers are connected to a segment. If the routers have similar routing tables, it may make sense to send only the default route (0.0.0.0) to hosts.

Figure 2-8. RIP routes to hosts