Efficient task scheduling in the cloud with queuing and multi-tactic Harris Hawks Optimization

Abstract

Cloud computing, a virtualization-based technology, faces challenges in task scheduling, which is crucial for cost-efficient execution and resource utilization. Ineffective VM Placement and shared physical machines lead to extended makespan, resource waste, energy consumption, increased inter-communication costs, and security breaches. This comprehensive approach aims to optimize VM allocation in cloud data centers, considering both energy efficiency and task dependencies. To decrease energy consumption and balance resource demands, the paper describes a method for assigning energy-efficient virtual machines (VMs) in cloud data centres that focuses on job dependencies and task execution times. The algorithm streamlines and increases efficiency by using queues for varying task intensities. By utilising an improved Multi-Tactic Harris Hawks Optimisation (MTHHO) algorithm, the suggested scheduling approach gets beyond drawbacks such local optima and poor convergence accuracy. Improved energy updating methods, elite opposition-based learning for flexibility, and Sobol sequences for population initialization are among the improvements. To avoid the process from getting trapped in local optima, the Gaussian walk learning technique is introduced. Therefore, the proposed heuristic-based MTHHO method balanced the load and allocate the resources effectively to improve QoS performances. The result shows that the proposed method of QoS performances attained less Makespan, energy consumption of 0.20, throughput of 2.4, and execution time of 16.75 with effectively allocated the resources of 98% when compared to the previous methods in cloud computing.