An Efficient Load Balancing Framework with Enhanced Parallel Processing and Scheduling Schemes for Heterogeneous Cloud Environment

Abstract

ABSTRACT newlineDynamic and agile Information Technology (IT) infrastructure creation newlineincludes the following capabilities: high-resource utilization, ability of demand newlineadjustment, provision of an immediate response to the diverse business conditions. newlineLow ownership cost and an effective resource management initiate the Virtual newlineMachine (VM) resizing and migrating strategies to satisfy the IT application newlinedemands. The main objectives of the adoption of virtualization concept are the newlineflexibility improvement and the ownership cost reduction. The interface construction newlineby the virtualized services allows the VM life cycle management activities with the newlinesharing capabilities. Virtualization mechanism extends interface applicability to newlineoffer the dynamic VM migration without disruption in activities. The leased newlineresources utilization in virtualization mechanism requires an effective scheduling to newlinesupport the parallel job processing with less time consumption. newlineThe complexity lies in the virtualized environment creation causes the newlineresource management and security challenges. The finite number of hosts with a newlinehuge workload and dynamic requirements according to the user needs are the newlinelimitations in virtualization. The allocation of Central Processing Unit (CPU) newlineproportional to the VM shares in scheduling provides the necessary time granularity newlinecompared to the other schedulers (E.g. fair share schedulers). The major requirement newlineof an effective virtualization and the CPU scheduling is the load management or newlinesharing. The large size job is divided into several small size tasks and allotted to the newlinetwo or more computers, CPUs, and network links are referred as load balancing that newlinereduces the delivery time and improves the efficiency. An effective resource newlinedistribution, minimum resource utilization, and the load imbalance are the major newlineproblems in load balancing strategy. This research work focuses on the effective newlineload balancing and scheduling in a cloud environment. newlineThe initial phase of the research work proposes the new approach called newlineOpti