Failover of VPN traffic between MPLS tunnels is accelerated by determining a main and a standby MPLS tunnel for each VPN and passing both the main and standby MPLS tunnel information to the forwarding plane of a network element configured to implement the VPN. In the event of failover of a main MPLS tunnel interface, the forwarding plane is instructed to use the standby MPLS tunnel instead of the main MPLS tunnel. Since the standby MPLS tunnel has already been determined, and information associated with the standby MPLS tunnel has already been provided to the forwarding plane, it is not necessary to recompute a new MPLS tunnel or reprogram the forwarding plane using the new MPLS tunnel information. By removing control plane operations and control plane-forwarding plane interaction from the handling of failovers, the failover recovery time may be accelerated, to thereby enable VPN connections to handle time sensitive traffic.