Experimental Study of Network Traffic Overhead in Cloud Environments
The continuous and rapid virtualization of computing has resulted in an explosion of cloud applications being utilized on a daily basis. The definition of cloud computing is fairly debated but can be summarized as the offloading of any computing resources to another remote machine. cloud computing allows for system resources, more specifically storage and computing power, available on demand without direct active management. With the development of cloud computing, it has benefited the technological scene massively over the past few years. With this massive benefit comes some interesting questions and issues regarding the performance and reliability of cloud computing architectures. By rigorously testing a virtual cloud environment, we will discuss open research challenges regarding the proposed framework. The testing included is primarily aimed to account for both collecting accurate network information and real-use case scenarios. The test cases will give insight into the capabilities of cloud environments in everyday usage scenarios. Cloud computing has a vast array of implementations as well as architectures, therefore it is critical to test a common cloud architecture and discover possible drawbacks and solutions for cloud computing environments. The cloud computing environment to be tested for the proposed test cases is an open source platform known as OpenStack. The overall performance of OpenStack is based on several components that make up OpenStack (Nova, Glance, Swift, Neutron, Cinder, Keystone, Horizon, etc.). The emphasis of the testing will be entirely based the networking component, Neutron, but Nova will be analyzed also to check computing statistics. The aim of this research is to open up discussion regarding possible performance solutions to be implemented in cloud computing data centers.
Grady, Adam Lee, "Experimental Study of Network Traffic Overhead in Cloud Environments" (2020). IdeaFest. 62.