Abstract
We consider data analytics workloads on distributed architectures, in particular clusters of commodity machines. To find a job partitioning that minimizes running time, a cost model, which we more accurately refer to as makespan model, is needed. In attempting to find the simplest possible, but sufficiently accurate, such model, we explore piecewise linear functions of input, output, and computational complexity. They are abstract in the sense that they capture fundamental algorithm properties, but do not require explicit modeling of system and implementation details such as the number of disk accesses. We show how the simplified functional structure can be exploited to reduce optimization cost. In the general case, we identify a lower bound that can be used for search-space pruning. For applications with homogeneous tasks, we further demonstrate how to directly integrate the model into the makespan optimization process, reducing search-space dimensionality and thus complexity by orders of magnitude. Experimental results provide evidence of good prediction quality and successful makespan optimization across a variety of operators and cluster architectures.
Similar content being viewed by others
References
Agarwal, R.C., Balle, S.M., Gustavson, F.G., Joshi, M., Palkar, P.: A three-dimensional approach to parallel matrix multiplication. IBM J. Res. Dev. 39(5), 575–582 (1995)
Akdere, M., Cetintemel, U., Riondato, M., Upfal, E., Zdonik, S.: Learning-based query performance modeling and prediction. In: ICDE, pp. 390–401 (2012)
Asanovic, K., Bodik, R., Demmel, J., Keaveny, T., Keutzer, K., Kubiatowicz, J., Morgan, N., Patterson, D., Sen, K., Wawrzynek, J., Wessel, D., Yelick, K.: A view of the parallel computing landscape. Commun. ACM 52(10), 56–67 (2009)
Ballard, G., Buluc, A., Demmel, J., Grigori, L., Lipshitz, B., Schwartz, O., Toledo, S.: Communication optimal parallel multiplication of sparse random matrices. In: SPAA, pp. 222–231 (2013)
Duggan, J., Cetintemel, U., Papaemmanouil, O., Upfal, E.: Performance prediction for concurrent database workloads. In: SIGMOD, pp. 337–348 (2011)
Duggan, J., Papaemmanouil, O., Çetintemel, U., Upfal, E.: Contender: a resource modeling approach for concurrent query performance prediction. In: EDBT, pp. 109–120 (2014)
Elmroth, E., Gustavson, F., Jonsson, I., K$\mathring{\text{a}}$gstr$\ddot{\text{ o }}$m, B.: Recursive blocked algorithms and hybrid data structures for dense matrix library software. SIAM Rev. 46(1), 3–45 (2004)
Ganapathi, A., Kuno, H.A., Dayal, U., Wiener, J.L., Fox, A., Jordan, M.I., Patterson, D.A.: Predicting multiple metrics for queries: Better decisions enabled by machine learning. In: ICDE, pp. 592–603 (2009)
Goodrich, M.T., Sitchinava, N., Zhang, Q.: Sorting, searching, and simulation in the mapreduce framework. In: ISAAC, pp. 374–383 (2011)
Gounaris, A., Kougka, G., Tous, R., Montes, C.T., Torres, J.: Dynamic configuration of partitioning in spark applications. IEEE Trans. Parallel Distrib. Syst. 28(7), 1891–1904 (2017)
Gunther, N.J., Puglia, P., Tomasette, K.: Hadoop superlinear scalability. Commun. ACM 58(4), 46–55 (2015)
Hahn, C., Warren, S., Eastman, R.: Extended edited synoptic cloud reports from ships and land stations over the globe, 1952–2009 (ndp-026c) (2012)
Herodotou, H., Babu, S.: Profiling, what-if analysis, and cost-based optimization of mapreduce programs. VLDB 4(11), 1111–1122 (2011)
Huang, B., Babu, S., Yang, J.: Cumulon: optimizing statistical data analysis in the cloud. In: Proceedings of SIGMOD, pp. 1–12 (2013)
Irony, D., Toledo, S., Tiskin, A.: Communication lower bounds for distributed-memory matrix multiplication. J. Parallel Distrib. Comput. 64(9), 1017–1026 (2004)
Kaoudi, Z., Quiane-Ruiz, JA., Thirumuruganathan, S., Chawla, S., Agrawal, D.: A cost-based optimizer for gradient descent optimization. In: SIGMOD, pp. 977–992 (2017)
Karloff, H., Suri, S., Vassilvitskii, S.: A model of computation for mapreduce. In: SODA, pp. 938–948 (2010)
Li, J., Naughton, J.F., Nehme, R.V.: Resource bricolage and resource selection for parallel database systems. VLDB J. 26(1), 31–54 (2017)
Li, R., Riedewald, M., Deng, X.: Submodularity of distributed join computation. In: (Upcoming) SIGMOD (2018)
Lichman, M.: UCI machine learning repository (2013)
Morton, K., Balazinska, M., Grossman, D.: Paratimer: a progress indicator for mapreduce dags. In: SIGMOD, pp. 507–518 (2010)
Munson, A.M., Webb, K., Sheldon, D., Fink, D., Hochachka, W.M., Iliff, M., Riedewald, M., Sorokina, D., Sullivan, B., Wood, C., Kelling, S.: The Ebird Reference Dataset, Version 2014. Cornell Lab of Ornithology and National Audubon Society, Ithaca, NY (2014)
Quinlan, J.R., et al.: Learning with continuous classes. Aust. Jt. Conf. Artif. Intell. 92, 343–348 (1992)
Ramakrishnan, R., Gehrke, J.: Database Management Systems, 3rd edn. McGraw-Hill, New York (2003)
Shi, J., Zou, J., Lu, J., Cao, Z., Li, S., Wang, C.: Mrtuner: a toolkit to enable holistic optimization for mapreduce jobs. In: VLDB, pp. 1319–1330 (2014)
Solomonik, E., Demmel, J.: Communication-optimal parallel 2.5d matrix multiplication and LU factorization algorithms. In: Euro-Par 2011 Parallel Processing, pp. 90–109 (2011)
Valsalam, V., Skjellum, A.: A framework for high-performance matrix multiplication based on hierarchical abstractions, algorithms and optimized low-level kernels. Concurr. Comput. 14(10), 805–839 (2002)
van de Geijn, R.A., Watts, J.: Summa: Scalable Universal Matrix Multiplication Algorithm. University of Texas at Austin, Tech. rep. (1995)
Venkataraman, S., Yang, Z., Franklin, M., Recht, B., Stoica, I.: Ernest: efficient performance prediction for large-scale advanced analytics. In: NSDI, pp. 363–378 (2016)
Vieth, E.: Fitting piecewise linear regression functions to biological responses. J. Appl. Physiol. 67(1), 390–396 (1989)
Wang, G., Chan, CY.: Multi-query optimization in mapreduce framework. In: VLDB, pp. 145–156 (2013)
White, B., Lepreau, J., Stoller, L., Ricci, R., Guruprasad, S., Newbold, M., Hibler, M., Barb, C, Joglekar, A.: An integrated experimental environment for distributed systems and networks. In: OSDI, pp. 255–270 (2002)
Wu, W., Chi, Y., Hacígümüş, H., Naughton, J.F.: Towards predicting query execution time for concurrent and dynamic database workloads. VLDB 6(10), 925–936 (2013)
Zhang, X., Chen, L., Wang, M.: Efficient multi-way theta-join processing using mapreduce. In: VLDB, pp. 1184–1195 (2012)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was supported by a Northeastern University (NU) Tier 1 award, by the National Institutes of Health (NIH) under Award Number R01 NS091421, by the Air Force Office of Scientific Research (AFOSR) under Grant Number FA9550-14-1-0160, and by the NSF Career Award under Award Number 1741634. The content is solely the responsibility of the authors and does not necessarily represent the official views of NU, NIH, AFOSR or NSF.
Rights and permissions
About this article
Cite this article
Li, R., Mi, N., Riedewald, M. et al. Abstract cost models for distributed data-intensive computations. Distrib Parallel Databases 37, 411–439 (2019). https://doi.org/10.1007/s10619-018-7244-2
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10619-018-7244-2