Title: SLA-aware task allocation with resource optimisation on cloud environment

Authors: Shrabanee Swagatika; Amiya Kumar Rath

Addresses: Department of Computer Science and Engineering, SOA Deemed to be University, Bhubaneswar, India ' Department of Computer Science and Engineering, VSS University of Technology, India

Abstract: Cloud computing is a materialised environment for several users due to its unpredictable nature. However cloud is a challengeable environment for many researches because of its hardness on predicting the accurate status of virtual machines (VM) in order to allocate resources for cloud consumers. Many researchers focus on optimisation and scheduling algorithms for predicting accurate status about VMs and their researches fail in providing the predominant results. In order to furnish efficacious results, we propose a novel cloud framework with resource optimisation. Two special components are chosen on it: one is resource manager (RM) and the other one is virtual machine re-allocator (VMRA). This framework has several methodologies like task clustering, task allocation, VM clustering and re-allocation of VMs. Specifically we introduce SLA-priority clustering algorithm for several tasks available on cloud, and these are allocated using ETC matrix with PSO algorithm. Our proposed cloud framework can be implemented using a CloudSim simulation tool which furnishes the efficient results. Thus our process effectively allocates resources on cloud environment and we also solve the problem of re-allocation of VM, power efficient VM re-allocation algorithm is proposed to solving the random failure issues.

Keywords: cloud computing resource optimisation; resource allocation; resource management; SLA constraints; particle swarm optimisation; PSO scheduling.

DOI: 10.1504/IJCNDS.2019.097651

International Journal of Communication Networks and Distributed Systems, 2019 Vol.22 No.2, pp.150 - 169

Received: 27 Jan 2018
Accepted: 19 Feb 2018

Published online: 04 Feb 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article