OSPF’s default dead timer is 40 seconds — meaning a failed link or neighbour can go undetected for nearly a minute before OSPF reconverges. BFD (Bidirectional Forwarding Detection) solves this by running lightweight hello packets at a much faster rate, pulling down OSPF adjacencies in under a second when a failure is detected. This video covers how to configure BFD for OSPF on Juniper logical systems.

BFD is configured at the per-interface level within the OSPF area hierarchy, not as a separate protocol. Under set protocols ospf area 0 interface, adding the bfd-liveness-detection option exposes two key parameters — minimum-interval, which sets how frequently BFD packets are sent in milliseconds, and multiplier, which sets how many missed packets trigger a failure declaration.

In the demonstration we configure BFD on the PE1 to PE3 OSPF link with a minimum interval of 300 milliseconds and a multiplier of 3. This gives us failure detection in 900 milliseconds — sub-second convergence compared to the 40-second dead timer. The same configuration is applied to both ends of the link, which is required for BFD to establish a session.

Once committed we verify the BFD session using show bfd session on PE3, confirming the session is up and transmitting at the configured interval. The output shows the neighbour address, session state, and the actual transmission rate — 300 milliseconds as configured.

The video explains the behaviour clearly — if three consecutive BFD packets are missed in either direction, the BFD session is declared down and OSPF immediately tears down the adjacency without waiting for the dead timer. Routes reconverge in under a second rather than waiting 40 seconds for OSPF to detect the failure on its own. The same BFD configuration syntax works across OSPF, ISIS and BGP, so this configuration approach carries over directly to other protocols in the series.