Backup Routers Overview
The purpose of the backup router is not to forward transit traffic. You (the system administrator) can use the backup router for local management of the routing device by way of the out-of-band management interface (fxp0, em0, or me0).
The routing protocol process (rpd) is the software process responsible for establishing routes. The rpd does not run while a routing device is booting. Therefore, the device has no routes. The backup router allows the routing device to install a route to a management network before the rpd is up and running.
You can use a backup router during the initial boot process, before any routing protocols have converged. The backup router allows the device to establish a Layer 3 connection quickly, which keeps management unavailability to a minimum. You can also use the backup router to manage the device if the routing protocol process fails to start properly.
The backup router connects directly to the local routing device (that is, the device connected on the same subnet) through its private management interface. To select a backup router, you commonly choose the default gateway of the management network that is directly connected to your routing device. Make sure that the specified backup router address is reachable and directly connected to the management interface.
Router A can be the backup router for Router B, and Router B can be the backup router for Router A, if the management interface of each router is connected to an interface on the other router, thus providing the necessary reachability.
When the rpd starts, the backup route (the route created by the backup router) is removed, and any default, static, or protocol-learned routes are installed.
Backup-router configurations can alter the backup Routing Engine kernel forwarding table even with nonstop active (NSR) enabled.