IP Routing Configuration

Setting Up the Initial IP Routing Table

Devices communicate with each other over routes. A route is a path from the sending device to the receiving device. Devices on a network learn about routes in a variety of ways. Routes can be manually configured by an administrator; devices can send out probes to discover how to get to a destination; or devices can receive updates about what routes are available. Once a device obtains information about a route, the device stores the route information in a routing table for future reference, as shown in Figure 10-1.

If the destination device is on the same network as the sending device, the sending device simply transmits the datagram directly to the destination. When a destination is not on the local network, a sending device forwards the datagram to a router. In order to forward a datagram, the sending device must first know what routers are connected to the local network.

Based on the network shown in Figure 10-1, a packet destined for 10.1.3.0 would be dropped. When a router first comes up, it knows only about the networks that are directly connected to its interfaces.

A router refers to entries about networks or subnets on which the router is directly connected. Each router connection to a network is configured with an IP address and mask, which are then associated with a router interface. The Cisco IOS software learns about this IP address and mask information from configuration information input from some source, such as a network administrator.

Routers learn about nonlocal routers and the shortest path to destination devices through a variety of methods, discussed next.

Understanding How IP Routing Learns Destinations

Routers learn paths to destinations in three ways:

• Static routes—Manually defined by the system administrator as the only path to the destination; useful for controlling security and reducing traffic.

• Default routes—Manually defined by the system administrator as the path to take when no route to the destination is known.

• Dynamically learned routes—Router learns of paths to destinations by receiving periodic updates from other routers.

IP routing is automatically enabled in the Cisco IOS software. To disable IP routing, enter the following command in global configuration mode:

Router(config)#no ip routing

When IP routing is disabled, the router will act as an IP end host for IP packets destined for or sourced by the router. To re-enable IP routing, issue the following command in global configuration mode:

Router(config)#ip routing

Note that this book primarily focuses on dynamic routing. Refer to the Cisco Press title Advanced Cisco Router Configuration for more information on static and default routes.

Specifying Administrative Distance Values

An administrative distance is a rating of the trustworthiness of a routing information source, such as an individual router or a group of routers. An administrative distance is an integer from 0 to 255. In general, the higher the value, the lower the trust rating. An administrative distance of 255 means the routing information source cannot be trusted at all and should be ignored.

Specifying administrative distance values enables the Cisco IOS software to discriminate between sources of routing information, as shown in Figure 10-2 To get to network 128.10.0.0, Router A will choose to send the packet to Router B because Router B has a lower administrative distance than Router C.

The software always picks the route whose routing protocol has the lowest administrative distance. Table 10-1 shows the default administrative distances for some routing information sources.

If a router has both routing protocols IGRP and RIP enabled, the Cisco IOS software uses the IGRP-derived information because the default IGRP administrative distance is lower than that for RIP and ignores the RIP-derived information. However, if you lose the source of the IGRP-derived information (for example, because of a power shutdown), the software uses the RIP-derived information until the IGRP-derived information reappears.

Configuring Static Routes

Static routes are user-defined routes that cause packets moving between a source and a destination to take a specified path. Static routes are important when the Cisco IOS software cannot build a route to a particular destination. Routers can forward packets only to known routes, and if the router cannot learn of a route dynamically, the static entry can be used to enable the router to route the incoming packet. Static routes are also useful for specifying a gateway of last resort to which all unroutable packets will be sent. This gateway (which is actually a router) is used as a last attempt to find some device to handle the packet.

Dynamic routing is typically preferred because static routing can be unwieldy in a large, complex, or volatile network because the administrator would have to make many manual changes. In small, simple, stable networks, however, static routing affords precision and control over the network without too much work.

To configure a static route, enter the ip route command in global configuration mode. A static route allows manual configuration of the routing table. No dynamic changes to this table entry will occur as long as the path is active. The complete parameters for the ip route command are as follows:

ip route network [mask] {address|interface} [distance] [permanent]

where the parameters have the following meanings:

• network—Destination network or subnet

• mask—Subnet mask

• address—IP address of next-hop router

• interface—Name of interface to use to get to destination network

• distance—The administrative distance

• permanent (Optional)—Specifies that the route will not be removed, even if the interface shuts down

If the mask is omitted in the ip route command, the router assumes it can use the default mask. Figure 10-3 provides a static route example based on the following ip route command:

Router(config)#ip route 172.16.1.0 255.255.255.0 172.16.2.1

In the example in Figure 10-3, the ip route command identifies the static route command; 172.16.1.0 specifies a static route to the destination subnetwork; 255.255.255.0 indicates the subnet mask (eight bits of subnetting are in effect); and 172.16.2.1 is the IP address of next-hop router in the path to the destination.

The assignment of a static route to reach the stub network 172.16.1.0 is proper for the Cisco A router because there is only one way to reach that network. A stub network is one that has only one connection to another network. If the network connects to more than one network and allows traffic to cross it to get from one network to another, that network is called a transit network.

The assignment of a static route from Cisco B to the cloud networks is also possible. However, a static route assignment is required for each destination network, so a default route may be more appropriate.

Because static routes have a low default administrative distance, they are always chosen over dynamic routes. You can change this effect by overwriting the administrative distance—essentially creating a static backup route—that is effective only when the protocol is down.

Configuring Default Routers

A router might not know the routes to all other networks. To provide complete routing capability, the common practice is to use some routers as default routers and give the remaining routers default routes to those routers.

To establish a default router, issue the following command in global configuration mode:

Router(config)#ip default-network network-number

where network-number is equal to the IP network number or subnet number defined as the default.

When an entry for the destination network does not exist in the routing table, the packet is sent to the default network, so the default network must exist in the routing table. One benefit of default routes is that they reduce the length of routing tables.

Use the default network number when you need a route but have only partial information about the destination network. Because the router does not have complete knowledge about all destination networks, it can use a default network number to indicate the direction to take for unknown network numbers.

In addition to the normal IP network addresses, IP Routing Information Protocol (RIP) uses 0.0.0.0 as the default route.

In the example shown in Figure 10-4, the ip default-network 192.168.17.0 global command defines the Class B network 192.168.17.0 as the destination path for packets that have no routing table entry.

To prevent unwanted updates from entering from the public network, company X could install a firewall in router A. To group those networks that will share company X's routing strategy, router A could implement an autonomous system number.

Grouping into Autonomous Systems

In Figure 10-4, you saw how company X used a default router to connect to a public network. It was mentioned that you could group routers into autonomous systems. An autonomous system is a set of routers and networks under the same administration. An autonomous system may consist of one router directly connected to one LAN to the Internet; or an autonomous system may be a corporate network linking several local networks through a corporate backbone. The autonomous system presents a consistent view of routing to the external world. For a router to belong to an autonomous system, all routers in that system must be:

• Interconnected

• Running the same routing protocol

• Assigned the same autonomous system number

The Network Information Center (InterNIC) assigns a unique autonomous system to enterprises. This autonomous system is a 16-bit number. A routing protocol such as Cisco's Interior Gateway Routing Protocol (IGRP) requires that you specify this unique, assigned autonomous system number in your configuration. A InterNIC-assigned autonomous system number is needed only if your organization plans to use an exterior router protocol, such as Border Gateway Protocol (BGP). If your company performs only interior routing, you need only ensure consistency and uniqueness of autonomous system numbers within your organization.

Using Interior or Exterior Routing Protocols

The design criteria for an interior routing protocol require it to find the best path through the network. In other words, the metric and how that metric is used is the most important element in an interior routing protocol.

Exterior protocols are used to exchange routing information between networks that do not share a common administration. IP exterior gateway protocols require the following three sets of information before routing can begin:

• A list of neighbor (or peer) routers or access servers with which to exchange routing information

• A list of networks to advertise as directly reachable

• The autonomous system number of the local router

As shown in Figure 10-5, the supported exterior gateway protocols are as follows:

• Border Gateway Protocol (BGP)

• Exterior Gateway Protocol (EGP)

An exterior routing protocol must isolate autonomous systems. Basically, another autonomous system is managed by some other staff. Because you have no control over how that network is configured, you need to protect the network against errors that could arise from misconfiguration. BGP and EGP are covered in more detail in the Cisco Press title Advanced Cisco Router Configuration.

Completing the IP Routing Configuration Tasks

The selection of IP as a routing protocol involves the setting of both global and interface parameters. Global tasks include:

• Select a routing protocol: RIP or IGRP (see Figure 10-7).

• Assign IP network numbers without specifying subnet values.

The interface task is to assign interface-specific addresses and the appropriate subnet mask.

Dynamic routing uses broadcasts and multicasts to communicate with other routers. The routing metric helps routers find the best path to each network or subnet.

Configuring Dynamic Routing

Two primary commands are used to configure dynamic routing: router and network. The router command starts a routing process; its form is as follows:

Router(config)#router protocol [keyword]

where the parameters specify the following:

• protocol—Either RIP, IGRP, OSPF, or Enhanced IGRP

• keyword—For example, autonomous system number, which is used with those protocols that require an autonomous system, such as IGRP

The network command is required because it allows the routing process to determine which interfaces will participate in the sending and receiving of routing updates. The network command starts up the routing protocol on all interfaces that the router has in the specified network. The network command also allows the router to advertise that network. Its form is as follows:

Router(config-router)#network network-number

where network-number specifies a directly connected network.

The network number must be based on the InterNIC network numbers, not subnet numbers or individual addresses. The network number also must identify a network to which the router is physically connected.