Configuring RIP

In keeping with the simple nature of RIP, configuration is an easy task. There is one command to enable the RIP process and one command for each network on which RIP is to run. Beyond that, RIP has few configuration options.

Case Study: A Basic RIP Configuration

Only two steps are necessary to configure RIP:

1. Enable RIP with the command router rip.

2. Specify each major network on which to run RIP with the net work command.

Figure 5.8 shows a four-router internetwork, with four major network numbers . Router Goober is attached to two subnets of network 172.17.0.0. The commands necessary to enable RIP are:

 Goober(config)#  router rip  Goober(config-router)#  network 172.17.0.0  

Similarly, Opie has two subnets of the same network and will configure with the following commands:

 Opie(config)#  router rip  Opie(config-router)#  network 172.17.0.0  

Figure 5.8. Both Andy and Barney are border routers between class-level networks.

Executing any router command puts the router into config-router mode, indicated in the prompt. The classful nature of RIP and the subnet hiding at network boundaries mean that no subnets can be specified with the network command ”only major class A, B, or C network addresses. RIP can run on any interface configured with any address belonging to the network specified with the network command.

Barney is attached to two networks ”10.0.0.0 and 192.168.83.0. Therefore, both networks must be specified:

 
 Barney(config)#  router rip  
 Barney(config-router)#  network  10.0.0.0  
 Barney(config-router)#  network  192.168.83.0  
 

Andy has one attachment to network 192.168.83.0, attachments to two subnets of 192.168.12.0, and attachments to two subnets of 172.17.0.0. Its configuration is:

 
 Andy(config)#  router rip  
 Andy(config-router)#  network 172.17.0.0  
 Andy(config-router)#  network 192.168.12.0  
 Andy(config-router)#  network 192.168.83.0  
 

In Figure 5.9, the command debug ip rip has been turned on in Andy. Of particular interest here is the subnet hiding that the router is performing. The subnets 192.168.12.64 and 192.168.12.192 are advertised between interfaces E0 and E2, which are both attached to network 192.168.12.0, but the network is summarized out E1, S0, and S1, which are all attached to different networks. Likewise, networks 192.168.83.0 and 172.17.0.0 are being summarized across classful boundaries. Notice also that Andy is receiving a summary route for network 10.0.0.0 from Barney. Finally, split horizon can be observed here. For example, the advertisement to Barney out E1 contains no entries for 10.0.0.0 or 192.168.83.0.

Figure 5.9. These debug messages show the RIP updates received and sent by router Andy. The results of both network summarization and split horizon can be observed in the update entries.

Case Study: Passive Interfaces

The router Floyd has been added to the internetwork (Figure 5.10). It is desired that no RIP advertisements be exchanged between Floyd and Andy. This is easy enough at Floyd:

 
 Floyd(config)#  router rip  
 Floyd(config-router)#  network 192.168.100.0  
 

Figure 5.10. Network policy calls for no RIP exchanges between Andy and Floyd.

By not including a network statement for 192.168.12.0, Floyd will not advertise on interface 192.168.12.66. Andy, however, has two interfaces attached to 172.17.0.0; the network must be included under RIP. To block RIP broadcasts on an interface connected to a subnet of a RIP-enabled network, add the passive-interface command to the RIP process. Andy's RIP configuration is:

  router rip   passive-interface Ethernet0   network 172.17.0.0   network 192.168.12.0   network 192.168.83.0  

Passive-interface is not a RIP-specific command; it may be configured under any IP routing protocol. Using the passive-interface command essentially makes a router a silent host on the data link specified. Like other silent hosts , it still listens to RIP broadcasts on the link and updates its routing table accordingly . If the desired result is to prevent the router from learning routes on the link, it must be achieved by more intricate control of routing updates, namely by filtering out updates. (Route filters are discussed in Chapter 13, "Route Filtering." ) Unlike a silent host, the router does not respond to a RIP Request received on a passive interface.

Case Study: Configuring Unicast Updates

Next , router Bea is added to the Ethernet link that Andy and Floyd share (Figure 5.11). The no-RIP policy between Andy and Floyd remains in place, but now Bea and Andy, as well as Bea and Floyd, must exchange RIP advertisements.

Figure 5.11. No RIP updates should be exchanged between Andy and Floyd, but both should exchange updates with Bea.

The configuration of Bea is straightforward:

  router rip   network 192.168.12.0   network 192.168.200.0  

The addition of a neighbor command under the RIP processes of Andy enables RIP to send a unicast advertisement to Bea's interface while the passive-interface command continues to prevent broadcast updates on the link. ^[9]

^[9] Another use of neighbor is to enable unicast updates on nonbroadcast media such as Frame Relay.

Andy's configuration is:

  router rip   passive-interface Ethernet0   network 172.17.0.0   network 192.168.12.0   network 192.168.83.0   neighbor 192.168.12.67  

Because Floyd must now send advertisements to Bea, a network command for 192.168.12.0 must be added. Passive-interface is also added to prevent broadcast updates, and a neighbor command is added to enable unicast updates to Bea:

  router rip   passive-interface Ethernet0   network 192.168.12.0   network 192.168.100.0   neighbor 192.168.12.67  

By enabling debug ip rip events at Andy, the results of the new configuration can be verified (Figure 5.12). Andy is receiving updates from Bea, but not from Floyd, and is sending updates directly to Bea's interface, but is not broadcasting on its E0.

Figure 5.12. The only updates Andy is sending out interface E0 are unicasts to Bea. Updates are being received from Bea, but not from Floyd.

Although Bea has learned routes from both Andy and from Floyd, and is broadcasting updates on the shared Ethernet, the policy still works because split horizon prevents Bea from advertising the routes learned from those two routers back onto the Ethernet.

Case Study: Discontiguous Subnets

In Figure 5.13, another router has been added to the internetwork with a subnet 10.33.32.0/20 on its E1 interface. The problem is that the other subnet of network 10.0.0.0, 10.33.0.0/20, is connected to Barney, and the only route between the subnets is via 192.168.83.0 and 192.168.12.0 ”two entirely different networks. As a result, network 10.0.0.0 is discontiguous.

Figure 5.13. Classful protocols such as RIP and IGRP cannot route a topology in which the subnets of network 10.0.0.0 are separated by different networks.

Barney will consider itself a border router between network 10.0.0.0 and network 192.168.83.0; likewise, Ernest_T will consider itself a border router between 10.0.0.0 and 192.168.12.0. Both will advertise a summary route of 10.0.0.0, and as a result Andy will be fooled into thinking that it has two equal-cost paths to the same network. Andy will load share on the links to Barney and Ernest_T, and there is now only a 50-50 chance that packets to network 10.0.0.0 will reach the correct subnet.

The solution is to configure subnets of network 10.0.0.0 on the same links on which 192.168.83.0/24 and 192.168.12.192/27 reside. This is accomplished with secondary IP addresses, as follows :

 Barney(config)#  interface e0  Barney(config-if)#  ip address 10.33.55.1 255.255.240.0 secondary  Andy(config)#  interface e1  Andy(config-if)#  ip address 10.33.55.2 255.255.240.0 secondary  Andy(config-if)#  interface e2  Andy(config-if)#  ip address 10.33.75.1 255.255.240.0 secondary  Andy(config-if)#  router rip  Andy(config-router)#  network 10.0.0.0  Ernest_T(config)#  interface e0  Ernest_T(config-if)  #ip address 10.33.75.2 255.255.240.0 secondary  

Because Andy did not previously have an interface on network 10.0.0.0, a network statement is added to the RIP process. The result of the configuration can be seen in Figure 5.14. The existing logical network structure remains in place, and a contiguous network 10.0.0.0 is "overlaid" onto it.

Figure 5.14. Secondary addresses are used to connect the subnets of network 10.0.0.0 across the same links on which other network addresses exist.

Figure 5.15 shows Ernest_T's routing table. Of interest here are the dual, equal-cost routes associated with next-hop addresses 10.33.75.1 and 192.168.12.195.

Figure 5.15. The routing process in this router sees the subnets 192.168.12.192/27 and 10.33.64.0/20 as separate links, although they reside on the same physical interface.

Note

The use of secondary addresses can contribute to congestion on the network.

Because the routing process sees secondary addresses as separate data links, caution should be used when planning them on RIP or IGRP networks. A separate RIP update will be broadcast on each subnet; if the updates are large and the bandwidth of the physical link is limited (such as on serial links), the multiple updates can cause congestion. Multiple updates on a link configured with secondary addresses can be observed in Figure 5.17, later in this chapter.

Type cautiously when entering secondary addresses. If you omit the keyword "secondary," the router will assume that the primary address is to be replaced with the new address. Making this mistake on an in-service interface will have serious consequences.

Case Study: Manipulating RIP Metrics

A serial link, to be used as a backup, has been added between Ernest_T and Barney (Figure 5.16). This link should be used only if the route via Andy fails. The problem is that the path between Barney's 10.33.0.0 subnet and Ernest_T's 10.33.32.0 subnet is 1 hop via the serial link and 2 hops via the preferred Ethernet links. Under normal circumstances, RIP will choose the serial link.

Figure 5.16. RIP metrics must be manipulated so that the 2-hop Ethernet route between Barney and Ernest_T will be preferred over the 1-hop serial route.

The route metrics can be manipulated with the offset-list command. The command specifies a number to add to the metric of a route entry and references an access list ^[10] to determine which route entries to modify. The syntax of the command is:

^[10] See Appendix B for a tutorial on access lists.

offset-list { access-list-number name } { in out } offset [ type number ]

The configuration of Ernest_T is:

 
 Ernest_T(config)#  access-list 1 permit 10.33.0.0 0.0.0.0  
 Ernest_T(config)#  router rip  
 Ernest_T(config-router)#  network 192.168.12.0  
 Ernest_T(config-router)#  network 10.0.0.0  
 Ernest_T(config-router)#  offset-list 1 in 2 Serial0  
 

An access list is written that identifies the route to subnet 10.33.0.0. The syntax of the offset list says, "Examine RIP advertisements incoming from interface S0. For route entries matching the addresses specified in access list 1, add 2 hops to the metric."

After Barney is configured, it will have the following entries in its configuration file:

  router rip   offset-list 5 in 2 Serial0   network 10.0.0.0   network 192.168.83.0   !   access-list 5 permit 10.33.32.0 0.0.0.0  

Figure 5.17 shows the results of the configuration at Ernest_T.

Figure 5.17. The addition of the hops specified in the offset list changes the hop count to subnet 10.33.0.0/20 via S0 from 1 to 3. Now the 2-hop route via E0 is used.

Alternatively, instead of having the two routers modify incoming routes, the routers can be configured to modify their outgoing routes. The following configurations will have the same effects as the previous configurations:

ERNEST_T:

  router rip   offset-list 3 out 2 Serial0   network 192.168.12.0   network 10.0.0.0   !   access-list 3 permit 10.33.32.0 0.0.0.0  

BARNEY :

  router rip   offset-list 7 out 2 Serial0   network 10.0.0.0   network 192.168.83.0   !   access-list 7 permit 10.33.0.0 0.0.0.0  

Several other options are available for configuring offset lists. If no interface is identified, the list will modify all incoming or outgoing updates specified by the access list on any interface. If no access list is called (by using a zero as the access list number), the offset list will modify all incoming or outgoing updates.

Note

Offset lists; incoming versus outgoing

Caution should be applied when choosing whether to use offset lists on incoming or outgoing advertisements. If more than two routers are attached to a broadcast network, consideration must be given to whether a single router should broadcast a modified advertisement to all its neighbors or whether a single router should modify a received advertisement.

Care must also be exercised when implementing offset lists on routes that are in use. When an offset list causes a next-hop router to advertise a higher metric than it had been advertising, the route will be marked unreachable until the holddown timer expires .