Green software defined distributed data center model

Abstract

With the advent of cloud computing and IoT, the demand for large capacity data centers are increasing rapidly. Large capacity datacenters are predominantly dependent on fossil fuel energy. Designing a distributed data center architecture with high availability, scalability and very small CO2 footprint is a challenge faced by the industry. The Software Defined Systems (SDSys) paradigm comes as a solution to provide reduced complexity of control and management in distributed data center designs. Software Defined Networking (SDN) particularly brings flexibility in network management and can be deployed to reduce energy consumption as well as better fault tolerance. The software-defined distributed green data center model provides a new approach to both the implementation of cloud data center services and the efficient use of green energy resources. This research is a comparative study on Distributed Software Defined Data Center (SDDC), by evaluating the traditional approach and proposing a method to reduce non-renewable energy consumption with better fault tolerance and robustness. In this report, we introduce a physically simulated software defined distributed data center model while examining opportunity to employ load balancing technique centrally to reduce brown energy consumption. We use OpenStack as the application and management layer along with OpenDaylight controller and Mininet emulator to simulate the model. We benchmark different performance metrics of the distributed system and compare them against standard.