An Enhanced Strategy to Minimize Makespan in Cloud Environment to Accelerate the Performance
With increase of dominance of Cloud Computing in Internet Communication Technologies, it becomes pertinent to have more number of techniques that are efficient and effective in improving system performance. Challenge here is efficient resource allocation
- PDF / 266,167 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 17 Downloads / 237 Views
Abstract With increase of dominance of Cloud Computing in Internet Communication Technologies, it becomes pertinent to have more number of techniques that are efficient and effective in improving system performance. Challenge here is efficient resource allocation to jobs requesting the resources in order to achieve customer satisfaction. Job Scheduling is a technique whereby the best schedule is constructed as a solution by applying various techniques and algorithms fulfilling the need of users so as they get their tasks completed in given time, cost, deadline or budget. This paper covers a new job scheduling algorithm Time Based Ant Colony Optimization Algorithm (TBACO) which is enthused from the basic Ant Colony Optimization algorithm. The objective is to minimize the makespan of given set of jobs, when they are scheduled on different datacenters which are the resource providers in the cloud. The new algorithm is modeled and simulated in Cloudsim simulator. Simulation results show the proposed TBACO algorithm outperformed basic ACO with 59 % savings in makespan. Keywords Job scheduling algorithms Virtual machines
Ant colony optimization Datacenters
H. Sachdeva (&) S. Kaushal A. Verma Department of Computer Science and Engineering, University Institute of Engineering and Technology, Panjab University, Chandigarh, India e-mail: [email protected] S. Kaushal e-mail: [email protected] A. Verma e-mail: [email protected] © Springer Science+Business Media Singapore 2016 S.C. Satapathy et al. (eds.), Proceedings of International Conference on ICT for Sustainable Development, Advances in Intelligent Systems and Computing 409, DOI 10.1007/978-981-10-0135-2_17
179
180
H. Sachdeva et al.
1 Introduction Cloud computing evolved from distributed computing and it provides scalable and dynamic on-demand resources. Grid computing can be part of cloud computing depending on user types. If the end users are systems managers and integrators, they install, upgrade virtualize servers and applications and overall take care of their maintenance in cloud. If the users are end customers, they need not to worry about how execution takes place cloud environment. It is a model for providing anytime, anywhere, opportune, scalable and on-demand system approach to a common pool of manageable computing resources, e.g., bandwidth, servers, storage, applications, and services that can be fastly provisioned and can also be released with least effort or service provider synergy [1]. This model consists of some important characteristics, service models, and four deployment models. The main characteristics are on-demand resource provisioning, measured service, high speed network access, resource sharing and pooling and rapid elasticity. The famous service models are Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS), Infrastructure-as-a-Service (IaaS), Anything as a Service (XaaS), etc. The main deployment models are Public Cloud, Private Cloud, Community Cloud and Hybrid Cloud.
2 Job Scheduling Job Scheduling is most bas
Data Loading...
