Static routes

Static routes are not automatically updated and must be manually reconfigured if the network topology changes. They can be redistributed into dynamic routing protocols, but dynamic routes cannot be redistributed into the static routing table.

Use static routes if your network has only one path to an outside network. You can also use static routes to control certain types of traffic or to secure specific links in a larger network. Most networks use dynamic routing protocols for communication, but you may configure one or two static routes for specific situations.

Advertisement of directly connected routes

Directly connected routes are advertised by IGP routing protocols when the interface is configured for the protocol.

If you create a static route that points to an interface, the system considers it directly connected. The routing protocol treats the destination as directly attached to the specified interface.

Route validity

A fully specified route is valid when the specified interface, IPv4 or IPv6, is enabled and up.

Recursive static route behavior and self-recursion

In a recursive static route, only the next hop is specified. The output interface is derived from the next hop. A recursive static route is valid only when the specified next hop resolves, either directly or indirectly, to a valid output interface, provided the route does not self-recurse and the recursion depth does not exceed the maximum IPv6 forwarding recursion depth.

• A route self-recurses if it is itself used to resolve its own next hop.

• If a static route becomes self-recursive, RIB sends a notification to static routes to withdraw the recursive route.

• It is not normally useful to manually configure a self-recursive static route, although it is not prohibited.

• A recursive static route that has been inserted in the routing table may become self-recursive as a result of some transient change in the network learned through a dynamic routing protocol.

• If this occurs, the static route is removed from the routing table, although not from the configuration. A subsequent network change may cause the static route to no longer be self-recursive, in which case it is re-inserted in the routing table.

Recursive static route example

This example shows how to specify that all destinations with address prefix 2001:0DB8::/32 are reachable through the host with address 2001:0DB8:3000::1:

Router# configure
Router(config)# router static 
Router(config-static)# address-family ipv6 unicast
Router(config-static-afi)# 2001:0DB8::/32 2001:0DB8:3000::1

Self-recursive static route example

A static route is self-recursive when the next hop depends on another route, which ultimately points to the original static route.

For example, assume:

• A BGP route: 2001:DB8:3000::/48 with next hop 2001:DB8::104

• A static route:

  2001:DB8::/32 → 2001:DB8:3000::1

Administrative distance and route preference

By default, the router uses static routes over dynamic routes because static routes have smaller administrative distances.

Use multicast static routes when multicast and unicast traffic require different network paths, such as when you need to create a tunnel because some network segments do not support multicast.

• Multicast static routes remain local on your router; they are not advertised or redistributed to other routers.

• You can use multicast static routes with GRE tunnels to bypass network segments that do not support multicast routing.

Key attributes of multicast static routes

Key attributes of multicast static routes include their local scope, use in RPF checks, and ability to separate multicast from unicast routing decisions.

• Local to the router

• Used for RPF checks

• Not advertised or redistributed

Note

Multicast static routes are not propagated to other routers and must be configured on each device as needed.

This figure illustrates the use of a GRE tunnel for multicast packets when the unicast path does not support multicast routing.