da Rosa Righi R, Andre da Costa C, Facco Rodrigues V, Cunha L (2017)
Publication Language: English
Publication Type: Journal article
Publication year: 2017
Pages Range: 1-18
DOI: 10.1504/ijcse.2017.10013365
Today, we observe that cloud infrastructures are gaining more and more space to execute HPC (High Performance Computing) applications. Unlike clusters and grids, the cloud offers elasticity, which refers to the ability of enlarging or reducing the number of resources (and consequently, processes) to support as close as possible the needs of a particular moment of the execution. In the best of our knowledge, current initiatives explore the elasticity and HPC duet by always handling the same number of resources at each scaling in or out of operation. This fixed elasticity grain commonly reveals a stair-shaped behaviour, where successive elasticity operations take place to address the load curve. In this context, this article presents GrainElastic: an elasticity model to execute HPC applications with the capacity to adapt the elasticity grain to the requirements of each elasticity operation. Its contribution concerns a mathematical formalism that uses historical execution traces and ARIMA time series model to predict the required number of resources (in our case, VMs) to address a reconfiguration point. Based on the proposed model, we developed a prototype that was compared with two other scenarios: (i) non-elastic application and (ii) elastic middleware with a fixed grain. The results presented gains up to 30% in favour of GrainElastic, showing us the relevance on adapting the elasticity grain to enhance system reactivity and performance.
APA:
da Rosa Righi, R., Andre da Costa, C., Facco Rodrigues, V., & Cunha, L. (2017). On providing on-the-fly resizing of the elasticity grain when executing HPC applications in the cloud. International Journal of Computational Science and Engineering, 1-18. https://dx.doi.org/10.1504/ijcse.2017.10013365
MLA:
da Rosa Righi, Rodrigo, et al. "On providing on-the-fly resizing of the elasticity grain when executing HPC applications in the cloud." International Journal of Computational Science and Engineering (2017): 1-18.
BibTeX: Download