Plinko

Abstract

This paper introduces Plinko, a network architecture that uses a novel forwarding model and routing algorithm to build networks with forwarding paths that, assuming arbitrarily large forwarding tables, are provably resilient against t link failures, ∀t ∈ N. However, in practice, there are clearly limits on the size of forwarding tables. Nonetheless, when constrained to hardware comparable to modern topof-rack (TOR) switches, Plinko scales with high resilience to networks with up to ten thousand hosts. Thus, as long as t or fewer links have failed, the only reason packets of any flow in a Plinko network will be dropped are congestion, packet corruption, and a partitioning of the network topology, and, even after t + 1 failures, most, if not all, flows may be unaffected. In addition, Plinko is topology independent, supports arbitrary paths for routing, provably bounds stretch, and does not require any additional computation during forwarding. To the best of our knowledge, Plinko is the first network to have all of these properties.

Fast ReRoute on Programmable Switches

“…Detecting port failures falls beyond the scope of this paper as it depends on specific hardware support. FRR mechanisms can be generally categorized along different dimensions: e.g., whether they tolerate only a single link/node failure [65], [66], [67] or multiple ones [68], [69]; whether routing tables are static (e.g., [11], [30], [60], [70], [69]) or dynamic (e.g., [71], [62]); whether packet header rewriting (e.g., [62], [61], [72], [68]) or packet duplication (e.g., [73]) is required; whether provide low stretch [74], [30] or maintain relatively low load [75], [76], [77].…”

Resilience support in software-defined networking: A survey

“…Nguyen et al [141] define algorithms for finding alternative paths if a link disruption occurs in the network; Stephens et al [144] present an architecture called Plinko, which is provably resilient to t link failures when the size of flow tables in the data plane is sufficiently large to accommodate backup flow rules; Borokhovich et al [147] use graph theory to model the network as a graph and run algorithms such as Depthfirst search (DFS) and Breadth-first search (BFS) to analyze the routing problem and ensure connectivity when a failure occurs.…”