Adaptive resilient routing via preorders in SDN

Abstract

In this paper, we propose and advocate a new routing paradigm-dubbed routing via preorders-which circumvents the limitations of conventional path-based routing schemes to effectively take advantage of topological diversity inherent in a network with rich topology for adaptive resilient routing, while at the same time meeting the quality-of-service requirements (e.g., latency) of applications or flows. We show how routing via preorders can be realized in SDN networks using the "match-action" data plane abstraction, with a preliminary implementation and evaluation of it in Mininet. CCS Concepts •Networks → Network protocol design; Routing protocols; Network design and planning algorithms;

Shortcutting Fast Failover Routes in the Data Plane

“…faster than in the control plane, yet still cannot avoid slow (quadratic) recovery times inherent to link reversal [8], and does not provide immediate protection as well. Ramadan et al [43] propose to speed up the convergence time by means of preorders and iteratively deactivating links. Lastly, Chiesa et al [13] proposed FRR primitives for programmable data planes which can complement ShortCut, and Stephens et al [54] investigate how to scale FRR rules by compression.…”

Shortcutting Fast Failover Routes in the Data Plane

“…DDC [36] implements link reversal algorithms in the data plane [29,57], and hence can recover from any non-partitioning failure set, faster than in the control plane, yet still cannot avoid slow (quadratic) recovery times inherent to link reversal [8], and does not provide immediate protection as well. Ramadan et al [44] propose to speed up the convergence time by means of preorders and iteratively deactivating links. Lastly, Chiesa et al [14] proposed FRR primitives for programmable data planes which can complement ShortCut, and Stephens et al [55] investigate how to scale FRR rules by compression.…”

When Raft Meets SDN

“…The control network (data plane) should not rely on the control plane to recover from failures, and should have pre-installed rules in switches for automatic failure recoveries. Fortunately, this can be achieved via a new routing paradigm proposed in [24], known as "routing via preorders", which provides adaptive resilient routing to ensure all-to-all communications among servers regardless of network failures, as long as the underlying network is not partitioned. We briefly discuss it as a prospective solution next.…”

“…The original PrOG is restored, when all links/nodes recover. PrOG also provides an additional feature, where the constructed graph can have a bounded threshold for the cost of the included paths from s to d (see [24] for more details.) Thus, a threshold can be defined for normal operations, and a relaxed threshold to be used upon failures.…”

“…In Section 4, we discuss possible "fixes" to circumvent these problems. In particular, we argue that in order to fundamentally break this inter-dependency, it is crucial to equip the SDN control network with a resilient routing mechanism such as PrOG [24] that guarantees connectivity among (non-partitioned) SDN controllers under arbitrary failures. Using a vanilla Raft implementation [19] and PrOG, we provide preliminary evaluation results in Section 5.…”