Cost-based meta-task scheduling algorithm for multicloud computing (CBMTSA)

Abstract

MultiCloud plays vital role in providing the heterogeneous types of resources to the user on-demand with minimum cost and time. Resource management and scheduling act as an influential aspect in the improvement of the performance of the MultiCloud environment. Scheduling deeds a considerable challenge in the distributed heterogeneous multiple cloud systems. The existing min-min algorithm is suitable for smaller number of tasks. The tasks are allocated to the VMs with high-processing speed. The proposed cost-based meta-task scheduling algorithm (CBMTSA) is feasible for the passive autonomous task-based scheduling for MultiCloud systems. Scheduling and rescheduling are two stages involved in this algorithm. Scheduling stage is used to allocate the tasks to the high-speed VMs. Makespan value is computed in the scheduling stage. The computed makespan is an outset value for rescheduling the tasks. The rescheduling stage is used to reschedule the tasks from high-speed VMs to low-speed VMs. The CBMTSA surpasses the existing min-min algorithm. Makespan, execution cost, and cloud server utilization ratio are the metrics considered in this algorithm.