Authors: Balamurugan Balusamy; Kotteswari Karthikeyan; Arun Kumar Sangaiah
Addresses: School of Information Technology and Engineering, VIT University, India ' Department of CSE, Annai Mira College of Engineering and Institute, Anna University, India ' School of Computing Science and Engineering, VIT University, India
Abstract: Scheduling a task and recovering resources in cloud computing is an important optimisation problem. Workload balancing of preemptive and non-preemptive task on the VM is a significant phase in task scheduling. The load of overloaded VM and under-loaded VM has to be balanced to achieve optimal machine utilisation. The recovery process is an essential phase when the failure in VM occurs. The resources that have been stored in the failed VM have to be recovered. The recovery process is done to reclaim the task that have been failed in the VM. In this paper, an algorithm named ant colony-based load balancing and fault recovery (ACB-LBR) is proposed that achieve well-balanced load across VM, activates recovery process at the time of VM failure and reduces power consumption among VMs. This algorithm uses for aging behaviour of Ant colony for balancing the tasks in overloaded VM that leads to high throughput. The ACB-LBR algorithm recovers the lost resource at failure time and manages less power consumption. The experimental results show that the proposed algorithm is effective compared to existing load balancing, recovery and power consumption algorithm.
Keywords: cloud computing; ant colony optimisation; ACO; load balancing; fault recovery; performance; metaheuristics; virtual machines; energy consumption.
International Journal of Advanced Intelligence Paradigms, 2017 Vol.9 No.2/3, pp.204 - 219
Received: 09 Jan 2016
Accepted: 10 Jul 2016
Published online: 17 Mar 2017 *