Workload Consolidation Using Task Scheduling Strategy Based on Genetic Algorithm in Cloud Computing

Abstract

Offering “Computing as a utility” on pay per use plan, Cloud computing has emerged as a technology of ease and flexibility for thousands of users over last few years. Distribution of dynamic workload among available servers and efficient utilization of existing resources in datacenter is one of the major concerns in Cloud computing. The load balancing issue needs to take into consideration the utilization of servers, i.e. the resultant utilization should not exceed the preset upper limits to avoid service level agreement (SLA) violation and should not fall beneath stipulated lower limits to avoid keeping some servers in active use. Scheduling of workload is regarded as an optimization problem that considers many varying criterion such as dynamic environment, priority of incoming applications, their deadlines etc. to improve resource utilization and overall performance of Cloud computing. In this work, a Genetic Algorithm (GA) based novel load balancing mechanism is proposed. Though not done in this work, in future, we aim to compare performance of proposed algorithms with existing mechanisms such as first come first serve (FCFS), Round Robin (RR) and other search algorithms through simulations.