VCE: A New Personated Virtual Cluster Engine for Cluster Computing

Abstract

Visualization addresses the problem of making more efficient use of available computer resources. Recently, virtual clusters try either to virtualize a single physical cluster into multiple independent virtual clusters to provide a virtual server as a highly scalable and highly available server built on a cluster of real servers, or to share physical clusters to offer economies of scale and more effective use of resources by multiplexing. However, since programs running on a cluster demonstrate different types of requirements as their executions proceeds, such as support for intensive processing, security, and massive data communications, it is therefore quite unrealistic to assume that a statically configured cluster with a predetermined number of nodes with specific features and support can do any good for such programs. This paper presents a different usage for virtualization in the context of distributed computing using virtual clusters, called virtual cluster engine (VCE), that provides a computing environment that can be both statistically and dynamically (re)organized according to the needs and requirements of programs, so that they can achieve the best possible performance as they suit themselves, indeed within the constraints of available resources. Feasibility of the proposed architecture for VCE has been studied on an experimental platform using seven real machines, VMWare ESX, VMotion, and VMWare programming kit, and a number of virtual machines. On average, 20% improvement on response times under VCE was experienced.