Abstract
The size and complexity of leading-edge high performance computing (HPC) systems and their electrical and cooling facilities have been continuously increasing over the years, following the increase in both their computational power and heat generation. Operational data analysis for monitoring the overall HPC system health and operational behavior has become highly important for a reliable and stable long-term operation as well as for operational optimizations. Operational log data collected from the HPC system and its facility can be composed by a wide range of information measured and sampled over time from different kind of sensors, resulting multivariate time-series log data. In our introduced visual analytics method, the HPC log data is represented as third-order tensor (3D array) data with three axes corresponding to time, space, and measured values. By applying multiple dimensionality reduction steps, characteristic time and space can be identified and be interactively selected for assisting the understanding of the HPC system state and operational behavior.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Fujiwara, T., et al.: A visual analytics framework for reviewing multivariate time-series data with dimensionality reduction. IEEE Trans. Visual. ComputerGr. 27(2), 1601–1611 (2021)
Guo, H., Di, S., Gupta, R., Peterka, T., Cappello, F.: La VALSE: scalable log visualization for fault characterization in supercomputers. In: Proceedings of EGPGV, pp. 91–100 (2018)
Kimura, T., et al.: Spatio-temporal factorization of log data for understanding network events. In: IEEE INFOCOM 2014-IEEE Conference on Computer Communications, pp. 610–618. IEEE (2014)
Kolda, T.G., Bader, B.W.: Tensor decompositions and applications. SIAM Rev. 51(3), 455–500 (2009)
McInnes, L., Healy, J., Saul, N., Grossberger, L.: UMAP: uniform manifold approximation and projection. J. Open Source Softw. 3(29), 861 (2018)
Pearson, K.: LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh Dublin Philos. Mag. J. Sci. 2(11), 559–572 (1901)
Shilpika, Lusch, B., Emani, M., Vishwanath, V., Papka, M.E., Ma, K.L.: MELA: a visual analytics tool for studying multifidelity HPC system logs. In: Proceedings of DAAC, pp. 13–18 (2019)
Tucker, L.R.: Some mathematical notes on three-mode factor analysis. Psychometrika 31(3), 279–311 (1966)
Xu, P., Mei, H., Ren, L., Chen, W.: ViDX: visual diagnostics of assembly line performance in smart factories. IEEE Trans. Visual. Comput. Graph. 23(1), 291–300 (2017)
Acknowledgements
This work was partially supported by JSPS KAKENHI (Grand Number 20H04194)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Fujita, K., Sakamoto, N., Fujiwara, T., Nonaka, J., Tsukamoto, T. (2022). A Visual Analytics Method for Time-Series Log Data Using Multiple Dimensionality Reduction. In: Chang, BY., Choi, C. (eds) Methods and Applications for Modeling and Simulation of Complex Systems. AsiaSim 2021. Communications in Computer and Information Science, vol 1636. Springer, Singapore. https://doi.org/10.1007/978-981-19-6857-0_3
Download citation
DOI: https://doi.org/10.1007/978-981-19-6857-0_3
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-19-6856-3
Online ISBN: 978-981-19-6857-0
eBook Packages: Computer ScienceComputer Science (R0)