On dimensioning optical grids and the impact of scheduling
When deploying Grid infrastructure, the problem of dimensioning arises: how many servers to provide, where to place them, and which network to install for interconnecting server sites and users generating Grid jobs? In contrast to classical optical network design problems, it is typical of optical Grids that the destination of traffic (jobs) is not known beforehand. This leads to so-called anycast routing of jobs. For network dimensioning, this implies the absence of a clearly defined (source, destination)-based traffic matrix, since only the origin of Grid jobs (and their data) is known, but not their destination. The latter depends not only on the state of Grid resources, including network, storage, and computational resources, but also the Grid scheduling algorithm used. We present a phased solution approach to dimension all these resources, and use it to evaluate various scheduling algorithms in two European network case studies. Results show that the Grid scheduling algorithm has a substantial impact on the required network capacity. This capacity can be minimized by appropriately choosing a (reasonably small) number of server site locations: an optimal balance can be found, in between the single server site case requiring a lot of network traffic to this single location, and an overly fragmented distribution of server capacity over too many sites without much statistical multiplexing opportunities, and hence a relatively large probability of not finding free servers at nearby sites.
KeywordsOptical networks Grids Anycast Dimensioning ILP Simulation
Unable to display preview. Download preview PDF.
- 2.De Leenheer, M., et al.: Design and Control of Optical Grid Networks (Invited). In: Proceedings Broadnets 2007, pp. 107–115. Raleigh, NC, 10–14 Sep. 2007Google Scholar
- 3.Fahramand, F., et al.: A multi-layered approach to optical burst-switched based Grids. In: Proceedings of the 5th International Workshop on Optical Burst/Packet Switching (WOBS) Co-located with IEEE/CreateNet BROADNETS 2005, Boston, MA, USA, vol. 2, pp. 1050–1057, 3 Oct. 2005Google Scholar
- 5.Stevens, T., et al.: Anycast routing algorithms for effective job scheduling in optical grids. In: Proceedings 32nd European Conference on Optical Communication (ECOC 2006), Cannes, France, vol. 3, pp. 371–372, 24–28 Sep. 2006Google Scholar
- 11.Bley, A., Koch, T., Wessäly, R.: Large-scale hierarchical networks: how to compute an optimal architecture? In: Proceedings of the 11th International Telecommunication Network Strategy and Planning Symposium (Networks 2004), Vienna, Austria, 13–16 June 2004Google Scholar
- 13.Thysebaert, P., De Turck, F., Dhoedt, B., Demeester, P.: Using divisible load theory to dimension optical transport networks for Grid excess load handling. In: Proceedings of the International Conference on Autonomic and Autonomous Systems & International Conference on Networking and Systems (ICAS/ICNS 2005), Papeete, Tahiti, 23–28 Oct. 2005Google Scholar
- 14.De Leenheer, M., Develder, C., De Turck, F., Dhoedt, B., Demeester, P.: Erlang reduced load model for optical burst switched grids. In: Proceedings of the 3rd International Conference on Networking and Services (ICNS2007), Athens, Greece, 19–25 June 2007Google Scholar
- 15.Rosberg, Z., Vu, H.L., Zukerman, M., White, J.: Blocking probabilities of optical burst switching networks based on reduced load fixed point approximations. In: Proceedings of the 22nd Annual Joint Conference of the IEEE Computer and Communication Societies (Infocom 2003), San Francisco, CA, USA,vol. 3, pp. 2008–2018, 30 March–3 April 2003Google Scholar
- 16.Christodoulopoulos, K., Doulamis, N., Kokkinos, P., Varvarigos, E.: Quality of service scheduling of computation and communication resources in Grid networks. In: Grid Computing Research Progress. Nova publishers (2008)Google Scholar
- 17.Christodoulopoulos, K., Doulamis, N., Varvarigos, E.: Joint communication and computation task scheduling in Grids. In: Proceedings of the 8th IEEE International Symposium On Cluster Computing and the Grid (CCGRID 2008), pp. 17–24, 19–22 May 2008Google Scholar
- 18.MacQueen, J.B.: Some methods for classification and analysis of multivariate observations. In: Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and Probability, vol. 1, pp. 281–297. University of California Press, Berkeley (1967)Google Scholar
- 19.Christodoulopoulos, K., Varvarigos, E., Develder, C., De Leenheer, M., Dhoedt, B.: Job demand models for optical grid research. In: Proceedings of the 11th International IFIP TC6 Conference on Optical Network Design and Modeling (ONDM2007), Athens, Greece. Lecture Notes in Computer Science, vol. 4534, pp. 127–136 (2007). doi:10.1007/978-3-540-72731-6_15s
- 23.GÉANT 2 project, http://www.geant2.net