DocumentCode :
3002051
Title :
Designing Flexible Resource Rental Models for Implementing HPC-as-a-Service in Cloud
Author :
Han Zhao ; Xiaolin Li
Author_Institution :
Dept. of Comput. & Inf. Sci. & Eng., Univ. of Florida, Gainesville, FL, USA
fYear :
2012
fDate :
21-25 May 2012
Firstpage :
2550
Lastpage :
2553
Abstract :
Due to its ability of delivering virtually unlimited computing power as a metered service, the emerging cloud computing paradigm becomes an attractive option for scientific communities to experiment large-scale resource-demanding applications traditionally deployed in high performance computing (HPC) centers. Planning resource rental is difficult in cloud environment because various parties may have fundamentally different interests. To accommodate such diverse heterogeneity, designing flexible resource rental models for multiple parties in the cloud market becomes a challenging issue. In this study, we investigate the problem of designing flexible resource rental models for implementing HPC-as-a-Service in cloud market. First, from the perspective of resource customers, we present cost-effective resource rental planning models to better utilize on-demand and spot resources in cloud. Next, we approach the problem from the perspective of resource providers, and propose a novel service scheduling model for a multi-tenant cloud resource sharing platform. The remaining work to complete the PhD dissertation is then presented. The proposed research highlights flexibility for resource rental planning in the following two aspects. (1) Flexibility for resource customers ensures them to deploy elastic HPC applications taking most advantage of both fixed-price on-demand resources and dynamic spot resources. (2) Flexibility for resource providers accommodates differentiated service requirements from various customers and achieves efficient HPC service scheduling. We believe that the proposed work presented in this paper offers a novel means to address the need for flexible resource rental models in cloud, and sheds light on future research in the broad area of resource management in cloud computing.
Keywords :
cloud computing; resource allocation; scheduling; HPC service scheduling; HPC-as-a-service implementation; cloud computing paradigm; cloud environment; cloud market; complete dynamic spot resources; cost-effective resource rental planning models; fixed-price on-demand resources; flexible resource rental model design; high performance computing centers; large-scale resource demanding applications; multitenant cloud resource sharing platform; resource management; resource providers; scientific communities; service scheduling model; Biological system modeling; Cloud computing; Computational modeling; Optimization; Planning; Proposals; Resource management; Cloud Computing; High Performance Computing; Optimization; Resource Rental Planning; Service Scheduling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel and Distributed Processing Symposium Workshops & PhD Forum (IPDPSW), 2012 IEEE 26th International
Conference_Location :
Shanghai
Print_ISBN :
978-1-4673-0974-5
Type :
conf
DOI :
10.1109/IPDPSW.2012.324
Filename :
6270891
Link To Document :
بازگشت