Conclusion

The new services deployed on IP networks, such as voice, online gaming and television, are placing ever-increasing demands on the availability of IP networks.

A first method for improving availability is rerouting. In the event of a failure, this consists in rapidly re-establishing communications by bypassing the faulty element. This is done dynamically by routing protocols, and recently major efforts have been made on the various stages of convergence to speed up the process and reduce packet losses.

A second method is to use high-availability (HA) functions to mask failures affecting only the control plane, not the transfer plane. This secures the control plane, restarting routing processes while maintaining packet switching. It also avoids rerouting in the network, and thus the associated protocol exchanges and impacts on other routers...