Dynamic switch-controller association and control devolution for SDN systems

Abstract

Abstract-In software-defined networking (SDN), as data plane scale expands, scalability and reliability of the control plane have become major concerns. To mitigate such concerns, two kinds of solutions have been proposed separately. One is multicontroller architecture, i.e., a logically centralized control plane with physically distributed controllers. The other is control devolution, i.e., delegating control of some flows back to switches. Most of existing solutions adopt either static switch-controller association or static devolution, which may not adapt well to the traffic variation, leading to high communication costs between switches and controller, and high computation costs of switches. In this paper, we propose a novel scheme to jointly consider both solutions, i.e., we dynamically associate switches with controllers and dynamically devolve control of flows to switches. Our scheme is an efficient online algorithm that does not need the statistics of traffic flows. By adjusting a parameter, we can make a tradeoff between costs and queue backlogs. Theoretical analysis and extensive simulations show that our scheme yields much lower costs or latency compared to other schemes, as well as balanced loads among controllers.In the last decade, cloud computing has emerged as the most influential computing paradigm to enable on-demand service hosting and delivery. Despite its importance, efficient resource allocation and network management in data centers are still main challenges to cloud providers.Previous works have proposed a variety of solutions to related problems, such as ensemble routing [15], energy budgeting [6], workflow scheduling [10], virtual slice provisioning [14], VM placement [20], etc. Meanwhile, softwaredefined networking (SDN) provides an alternative perspective to manage the whole network. The key idea of SDN is to decouple the control plane from the data plane [12]. In such a way, data plane can focus on performing basic functionalities such as packet forwarding at high speed, while the logically centralized control plane manages the whole network. Usually, switches send requests to the control plane for processing some flow events, e.g., flow install events.The control plane is a potential bottleneck of SDN in terms of scalability and reliability. As the data plane expands, control plane may not be able to process the increasing number of requests if implemented with a single controller, resulting unacceptable latency to flow setup. Reliability is also an issue since a single controller is a single point of failure, which may result in the break-down of the control plane and the entire network.Existing proposals to address such problems fall broadly into two categories. One is to implement the control plane as a distributed system with multiple controllers [7] For switch-controller association, the first category of solution, the usual design choice is to make a static switchcontroller association [7] [17]. However, such static association may result in overloading of controllers and inc...

BFT Protocols for Heterogeneous Resource Allocations in Distributed SDN Control Plane

“…The controllers in the Internet2 scenario were placed so to maximize the system coverage against failures as per [6], [10]. The controllers of the fat-tree topology were placed on the leaf-nodes as per [6], [11]. The state synchronization between the controllers and the resulting switch reconfigurations occur in in-band control mode.…”

MORPH: An Adaptive Framework for Efficient and Byzantine Fault-Tolerant SDN Control Plane

“…Advanced Networking -https://www.internet2.edu/ products-services/advanced-networking/ 2 Open vSwitch -https://www.openvswitch.org/ (a) Internet2 topology [38] (b) Fat-tree topology [39] Fig. 4.…”

“…Elements highlighted in green and blue represent the switches and clients, respectively. Red elements are the controller instances placed as per [38], [39]. Each switch instance in the Internet2 topology is allocated a client, while the fat-tree topology hosts clients at the leaf switches.…”

“…The links of the fattree topology only posses the inherit processing and queuing delays. The arrival rates of the incoming service embedding requests were modeled using a negative exponential distribution [39]. In the fat-tree topology, each leaf-switch was connected to 2 client instances, bringing the total number of clients up to 16.…”

“…4a. The SDN controller replicas in the datacenter topology were deployed on the leaf-nodes, similar to the controller placement presented in [39]. To evaluate the effect of the number, ratio and order of the Byzantine and availability-type failures, we deploy a specialized fault injection component.…”