Dragonfly addressing model for software defined networks based on datacenters (original) (raw)

SDN-Based Load Balancing Scheme for Fat-Tree Data Center Networks

Al-Nahrain Journal for Engineering Sciences, 2017

This paper proposes a new load balancing algorithm for data center networks by means of exploiting the characteristics of Software Defined Networks. Mininet was utilized as an emulation tool for the purpose of emulating and evaluating the proposed design, Miniedit was utilized as a GUI tool for the same purpose. In order to obtain a realistic environment to the data center network, Fat-Tree topology was utilized with the following parameters; 4 pods, 16 edge switches, 16 aggregation switches, 4 core switches, and 16 hosts. Different scenarios and traffic distributions were applied in order to cover as much possible cases of the real traffic. POX controller was chosen as an SDN controller.The suggested design showed outperformance when compared to the traditional scheme in term of throughput and loss rate for all the evaluated scenarios. The first scenario assumes joining of new hosts while in the second scenario; there was an increase in the demand of the already established connect...

SOFTWARE DEFINED NETWORK

— As the network traffic increasing daily, the consumption of resources increasing with the same rate. Cloud computing data centers are becoming gradually popular and important for provisioning the computer networking resources. In the past, it has been observed that a lot of importance was given to the resources that could be employed using Virtualization technology. But recently, the focus is moving from the data center for LAN and WAN to Cloud computing as the core system infrastructure. The requirements for such implementations depend on computer resources, storage and networking which in a way is becoming Software Defined Network (SDN). Instead of limiting the need for physical infrastructure and applications, the precisely defined network in virtual environment is the future. Software Defined Network plays an important role in providing the common platform to all the possible applications, innovations, operators, vendors and hardware components. The main objective of this literature review is to do research work and describes the ongoing efforts to address the thought-provoking issues.

Software Defined Networking: an emerging paradigm to computer networking

2017

SOFTWARE DEFINED NETWORKING - An Emerging Paradigm To Computer Networking Software Defined Networking (SDN) is an emerging paradigm in networking technology that enables innovation on how network systems are managed and designed. SDN plays a huge role in the effort to make computer networks programmable. This thesis discusses the history and efforts to programmable and active networks, the early practices towards separating the control plane and data plane. Highlighting the architectural concepts in networking that software defined networks emanated as well as the history and evolution of software defined networking. The thesis further discusses how SDN simplifies the complexity of managing large and distributed network system. It also examines the technologies that support software defined networking such as Network Virtualization and OpenFlow. The thesis reviews the components of software defined network architecture: the data plane layer, the controller plane layer, and application layer. Furthermore, it reports the practical implementation of SDN in network functions virtualization. Finally, it depicts the notion of software defined networking, whose Application Programming Interface may be implemented by OpenFlow. The practical aspect of this thesis shows a simple emulation of software-defined network architecture using Mininet network emulation tool. Hopefully, this work aims to benefit those who intend to learn about the fundamental principles of Software Defined Networking.

A generalized resource allocation framework in support of multi-layer virtual network embedding based on SDN

Network Virtualization (NV) allows multiple heterogeneous architectures to simultaneously coexist in a shared infrastructure. Embedding multiple virtual networks (VNs) in a shared substrate deals with efficient mapping of virtual resources in the physical infrastructure and is referred to as the Virtual Network Embedding problem (VNE-problem). Although there is recently a number of research work in the area of network virtualization based on the SoftwareDefined Networking (SDN) technology, virtual network embedding in SDN remains challenging from both theoretical and practical points of view. This article focuses on virtual network embedding strategies and related issues for Infrastructure-as-a-Service (IaaS) paradigms under the constraint of fixed virtual node locations. Special considerations are given to the problems related to resource allocation and link sharing of multi-layer virtual networks on top of the physical substrate. Firstly, a heuristic virtual network embedding algorithm is proposed that can improve the mapping acceptance ratio and resource efficiency in the IaaS context. Secondly, REsource reSERvation in generalized Virtual NETworks (ReServNet), a Software-Defined Networking platform designed for embedding multi-level virtual networks in physical infrastructures is developed. By defining new softwarized logical functions, ReServNet allows network administrators to create and manage multiple virtual networks on top of the physical network and allocate bandwidth resources to them accordingly. Moreover, the ReServNet framework allows for designing, prototyping, benchmarking and evaluating the performance of different network embedding algorithms easily in real SDN virtualization environments. Different issues related to virtual network embedding on SDN-based physical substrate are also analyzed and discussed in detail.