Static and default routes

Why type routes by hand?

A router knows its connected networks automatically. To reach any other network it needs a route. You can either run a routing protocol, which learns routes by itself, or type static routes. A static route never changes on its own: it uses no CPU or bandwidth for updates and does exactly what you said. That is good for small, stable networks and bad for big, changing ones.

The basic command

R1(config)# ip route 192.168.20.0 255.255.255.0 10.0.12.2
R1(config)# ip route 192.168.20.0 255.255.255.0 GigabitEthernet0/1
R1(config)# ip route 192.168.20.0 255.255.255.0 GigabitEthernet0/1 10.0.12.2

The format is ip route, then the destination network, its subnet mask (not a prefix length), then where to send the packets.

• Next hop IP (first line): the most common form. The router does a second lookup to find which interface leads to 10.0.12.2. This is called a recursive lookup.
• Exit interface (second line): the route shows as directly connected. Fine on point-to-point serial links. On Ethernet it is a poor choice, because the router must ARP for every destination address and depends on the next router answering for them (proxy ARP).
• Fully specified (third line): both the interface and the next hop. No recursive lookup, and it works cleanly on Ethernet.

A static route goes into the routing table only if it is usable: the exit interface must be up, or the next hop must be reachable through some other route. Check with show ip route static.

Default route

R1(config)# ip route 0.0.0.0 0.0.0.0 203.0.113.1
R1# show ip route | include 0.0.0.0
Gateway of last resort is 203.0.113.1 to network 0.0.0.0
S*    0.0.0.0/0 [1/0] via 203.0.113.1

Small offices and branches usually need just one default route pointing at the ISP or the head office router. This is called a stub network: there is only one way out.

Host routes and floating static routes

A host route has mask 255.255.255.255 (/32) and matches exactly one address, for example a single server you want to reach over a special path: ip route 10.9.9.9 255.255.255.255 10.0.13.3.

A floating static route is a backup. You add an administrative distance at the end of the command, higher than the main route's AD. While the main route exists, the backup "floats" out of the table. When the main route disappears, the backup is installed.

! Main path through the leased line (AD 1, the default for static routes)
R1(config)# ip route 0.0.0.0 0.0.0.0 10.0.12.2
! Backup through broadband, AD 5: used only if the first route disappears
R1(config)# ip route 0.0.0.0 0.0.0.0 198.51.100.1 5

A static route with a next hop disappears only when the router can no longer reach that next hop. If the ISP fails beyond the modem while your interface stays up, the route stays and traffic is lost. Larger networks add tracking (Cisco IP SLA) to remove the route when a test ping fails; check your platform's documentation for it.

IPv6 static routes

R1(config)# ipv6 unicast-routing
R1(config)# ipv6 route 2001:db8:20::/64 2001:db8:0:12::2
R1(config)# ipv6 route ::/0 2001:db8:0:12::2
R1(config)# ipv6 route ::/0 GigabitEthernet0/1 fe80::2

Without ipv6 unicast-routing the router will not forward IPv6 packets at all. ::/0 is the IPv6 default route.

When static routing is good, and when it is not

Troubleshooting: the missing return route

The most common static routing problem: R1 has a route to the LAN behind R2, but R2 has no route back to the LAN behind R1. PC1's ping reaches PC2, but the reply dies at R2, so ping shows "Request timed out". Every conversation needs a route in both directions.

1. Ping step by step: the gateway first, then the next router, then the far LAN.
2. Run traceroute (tracert on Windows) to see the last router that answers.
3. On each router, run show ip route and check routes to BOTH the destination and the source network.
4. Check typing mistakes: a wrong mask, a wrong next hop, or a next hop that is not on a connected subnet.