Tuning Application in a Multi-cluster Environment
The joining of geographically distributed heterogeneous clusters of workstations through the Internet can be a simple and effective approach to speed up a parallel application execution. This paper describes a methodology to migrate a parallel application from a single-cluster to a collection of clusters, guaranteeing a minimum level of efficiency. This methodology is applied to a parallel scientific application to use three geographically scattered clusters located in Argentina, Brazil and Spain. Experimental results prove that the speedup and efficiency estimations provided by this methodology are more than 90% precision. Without the tuning process of the application a 45% of the maximum speedup is obtained whereas a 94% of that maximum speedup is attained when a tuning process is applied. In both cases efficiency is over 90%.
KeywordsLocal Cluster Network Throughput Maximum Speedup Parallel Application Tuning Process
Unable to display preview. Download preview PDF.
- 3.Javadi, B., Akbari, M., Abawajy, J.: Performance analysis of heterogeneous multi-cluster systems. In: ICPP 2005, pp. 493–500 (2005)Google Scholar
- 4.Bal, H.E., Plaat, A., Bakker, M.G., Dozy, P., Hofman, R.F.H.: Optimizing parallel applications for wide-area clusters. In: Proceedings of IPPS/SPDP 1998, pp. 784–790 (1998)Google Scholar
- 6.Aida, K., Natsume, W., Futakata, Y.: Distributed computing with hierarchical master-worker paradigm for parallel branch and bound algorithm. In: CCGrid 2003, pp. 156–163 (2003)Google Scholar
- 9.Carusela, M.F., Perazzo, R.P.J., Romanelli, L.: Stochastic resonant memory storage device. Physical Review 64(3 pt 1), 31101 (2001)Google Scholar
- 11.McNamara, B., Wiesenfeld, K.: Theory of stochastic resonance. Physical Review A 39, 4854–4869 (1989)Google Scholar