Dynamic VM Provisioning for TORQUE in a Cloud Environment

20th International Conference on Computing in High Energy and Nuclear Physics

物理学;计算机科学

Zhang, S.^1 ; Boland, L.^2 ; Coddington, P.^1 ; Sevior, M.^2

EResearch SA, Adelaide

SA, Australia^1

University of Melbourne, Melbourne

VIC, Australia^2

关键词: Cloud environments;    Common resources;    Computational infrastructure;    Computational requirements;    Distributed resources;    Educational sectors;    High-throughput computing;    Tools and techniques;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/513/3/032107/pdf
DOI  :  10.1088/1742-6596/513/3/032107

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Cloud computing, also known as an Infrastructure-as-a-Service (IaaS), is attracting more interest from the commercial and educational sectors as a way to provide cost-effective computational infrastructure. It is an ideal platform for researchers who must share common resources but need to be able to scale up to massive computational requirements for specific periods of time. This paper presents the tools and techniques developed to allow the open source TORQUE distributed resource manager and Maui cluster scheduler to dynamically integrate OpenStack cloud resources into existing high throughput computing clusters.