Avoiding bottlenecks in networks by short paths
- PDF / 890,703 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 42 Downloads / 233 Views
Avoiding bottlenecks in networks by short paths Chen Levi1 · Michael Segal1 Accepted: 14 September 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Data center networks are typically characterized by high density communication components that process and exchange large amount of information using shortest paths. Mostly, data centers network topologies contain multi-rooted tree with multiple equal cost shortest paths between pairs of hosts. Usually, data center networks operation is based on the result of shortest path algorithms and per-flow static hashing which may cause poor network utilization rates with some links becoming congested while some parts of the network are underused. This work presents a flow scheduling algorithm that exploits the path diversity in data center topologies and dynamically reroutes large flows through less congested shortest and non-shortest paths based on the current state of the network without causing packets reordering. The algorithm aims to minimize flows latency while maximizing the network utilization rates. Results show that the algorithm proposed in this work reduces flows completion time by 13–24% over Equal-Cost Multi-Path routing, while improving the average network utilization by up to 10%. Keywords Short paths routing · Data centers · Flow scheduling solution · Performance guarantee
1 Introduction Data centers contain large number of switches connected by a communication network and have to process and exchange large amounts of data in short time periods. In order to be able to meet communication requirements, data center networks topologies are typically densely connected and characterized by “multi-rooted” tree topologies with multiple core switches that contain multiple equal cost shortest paths between pairs of hosts [1]. Traditionally, data center networks are operated using commodity packet communication mechanisms that are based on the result of shortest path algorithms and traffic engineering using node-local optimization operations [2,10]. The lack of global knowledge about the network results in sub-optimal overall network operation and poor network utilization rates with some links becoming congested while some parts of the network are underused [4,5]. In recent years, Software Defined Networks (SDNs), which are enabled by communication protocols such as Open Flow [14], provide an alternative in this matter. It is possiThe work on this paper has been partially supported by Israeli Innovation Authority.
B 1
Michael Segal [email protected] Ben-Gurion University of the Negev, 84105 Beer-Sheva, Israel
ble to define new control plane protocols in software and to make use of centralized network information including network topology and available paths. SDN also offers a communication between the control plane and the data plane. It enables the control plane to take advantage of the centralized flow statistics in the network, to control forwarding tables of all switches and provides the controller information of current traffic di
Data Loading...
