Achieving convergence-free routing using failure-carrying packets

Abstract

Current distributed routing paradigms (such as link-state, distancevector, and path-vector) involve a convergence process consisting of an iterative exploration of intermediate routes triggered by certain events such as link failures. The convergence process increases router load, introduces outages and transient loops, and slows reaction to failures. We propose a new routing paradigm where the goal is not to reduce the convergence times but rather to eliminate the convergence process completely. To this end, we propose a technique called Failure-Carrying Packets (FCP) that allows data packets to autonomously discover a working path without requiring completely up-to-date state in routers. Our simulations, performed using real-world failure traces and Rocketfuel topologies, show that: (a) the overhead of FCP is very low, (b) unlike traditional link-state routing (such as OSPF), FCP can provide both low lossrate as well as low control overhead, (c) compared to prior work in backup path precomputations, FCP provides better routing guarantees under failures despite maintaining lesser state at the routers.

High Availability in the Future Internet

“…Some of them change the traditional IP's destination based forwarding [32,51,16,3], while others use signaling to indicate that a packet is on detour [22,11,1,49,31,33]. On the other hand, there are methods, which use explicit tunneling to avoid the failed components [7,17,36,6].…”

Virtual Area Routing: a Scalable Intra-Domain Routing Scheme

“…[3] Another approach to solve the scalability problem is to avoid the burst of control traffic. FCP [5] aims to eliminate convergence period by carrying failure information along the regular packets. Routers can compute an available path with failure information in time and no flooding exists.…”

Handling Multiple Failures in IP Networks through Localized On-Demand Link State Routing

“…Hence, it is important to devise schemes that protect the network against not only single failures but also multiple independent failures. Our work is motivated by this need, which is also the focus of some of the recently proposed routing schemes [5]- [7].…”

“…Recently, [7] proposed an approach to handle dual link, but only single node failures. On the other hand, failure carrying packets (FCP) [5] and packet recycle (PR) [6] try to forward packets to reachable destinations even in case of arbitrary number of failures. The drawbacks, however, are that FCP carries failure information in each packet all the way to the destination whereas PR forwards packets along long detours.…”