A New Approach on Network Resources Management in Grids
Currently, Grid applications are usually developed for a network offering only a best effort packet delivery service. Nevertheless, High Performance Computing applications with Quality of Service requirements stress the capabilities of the interconnecting network of the target Grid infrastructure. Therefore, the application runtime environment must interact, through a Grid Resource Broker, with a Network Resource Manager to obtain information about the network and to reserve network resources among the computational resources that will host the execution of the applications.
The paper presents the design and the development of a Grid Network-Aware Resource Broker. It enhances the features of a Grid Resource Broker with the capabilities provided by a Network Resource Manager. The innovative contribution of the presented integration is the possibility to design and implement new mapping/scheduling mechanisms to take into account both network and computational resources. Finally, we will show how to exploit the new features offered by the Grid Network-Aware Resource Broker in scheduling parallel applications with QoS requirements.
KeywordsGrid Computing High-level Programming Environment Resource Management Quality of Service MPLS DiffServ
Unable to display preview. Download preview PDF.
- M. Dìaz, B. Rubio, E. Soler, and J. M. Troya, “SBASCO: Skeleton-based scientific components,” in Proc. 12th Euromicro Conference on Parallel, Distributed, and Network-Based Processing (PDP’04), A Coruña, Spain, Feb. 2004.Google Scholar
- H. Kuchen, “A Skeleton Library,” in Proc. 8th Euro-Par Conference, London, UK, Aug. 2002.Google Scholar
- T. Ferrari, (ed). “Grid Network Services Use Cases, Version 2.10.”. Grid High Performance Networking Research Group, July 7, 2006.Google Scholar
- B. Lowekamp, B. Tierney, L. Cottrell, R. Hughes-Jones, T. Kielmann, M. Swany. “A Hierarchy of Network Performance Characteristics for Grid Application and Services,” in Grid Forum Document, GFD-R.023, Global Grid Forum, May 24, 2004.Google Scholar
- D. Adami, N. Parlotti, S. Giordano, M. Repeti. “Design and development of a GNRB for the coordinated use of network resources in a high performance grid environment”, in Distributed Cooperative Laboratories, pp.295 – 307, Springer-Verlag, 2005.Google Scholar
- E. Demaine and N. Immorlica. “Correlation Clustering with Partial Information”, in Proceedings of the 6th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, Berlin, Germany, 2003.Google Scholar
- Sinnen, O. and L. Sousa, “A Classification of Graph Theoretic Models for Parallel Computing”. Technical report, Instituto Superior Tecnico, Technical University of Lisbon, Portugal, 1999.Google Scholar
- R. Baraglia, M. Danelutto, T. Fagni, D. Laforenza, S. Orlando, A. Paccosi, N. Tonellotto, M. Vanneschi, and C. Zoccolo. “HPC application execution on Grids”, in V. Getov, D. Laforenza, and A Reinefeld, editors, Future Generation Grids. Springer-Verlag CoreGRID Series, 2005.Google Scholar
- Grid.it: Enabling Platforms for High-Performance Computational Grids Oriented to Scalable Virtual Organizations. http://grid.it/.Google Scholar