Scheduling Divisible Loads to Optimize the Computation Time and Cost
Abstract
Efficient load distribution plays an important role in grid and cloud applications. In a typical problem, a divisible load should be split into parts and allocated to several processors, with one processor responsible for the data transfer. Since processors have different speed and cost characteristics, selecting the processor order for the transmission and defining the chunk sizes affect the computation time and cost. We perform a systematic study of the model analysing the properties of Pareto optimal solutions. We demonstrate that the earlier research has a number of limitations. In particular, it is generally assumed that the load should be distributed so that all processors have equal completion times, while in fact this property is satisfied only for some deadlines; for many optimal schedules this property does not hold. Moreover, fixing the processor sequence in the non-decreasing order of the cost-characteristic may be appropriate only for Pareto-optimal solutions with relatively large deadlines; optimal schedules for tight deadlines may have a different order of processors. We conclude with an efficient algorithm for finding the time-cost trade-off.
Keywords
Scheduling Divisible Load Time/cost OptimizationPreview
Unable to display preview. Download preview PDF.
References
- 1.Abdullaha, M., Othman, M.: Cost-based multi-QoS job scheduling using divisible load theory in cloud computing. In: Proceedings of the 2013 International Conference on Computational Science (ICCS), pp. 928–935 (2013)Google Scholar
- 2.Beaumont, O., Legrand, A., Robert, Y.: Scheduling divisible workloads on heterogeneous platforms. Parallel Comput. 29, 1121–1152 (2003)MathSciNetCrossRefGoogle Scholar
- 3.Berlińska, J., Drozdowski, M.: Scheduling divisible MapReduce computations, J. Parallel and Distrib. Comput. 71, 450–459 (2011)CrossRefGoogle Scholar
- 4.Buyya, R., Abramson, D., Venugopal, S.: The grid economy. Proceedings of the IEEE 93, 698–714 (2005)CrossRefGoogle Scholar
- 5.Charcranoon, S., Robertazzi, G.R., Luryu, S.: Parallel processor configuration design with processing/transmission costs. IEEE Trans. on Computers 49, 987–991 (2000)CrossRefGoogle Scholar
- 6.Choi, K., Robertazzi, T.G.: Cost performance analysis in multi-level tree networks. In: Proceedings of the Ninth International Symposium on Parallel and Distributed Computing (ISPDC), pp. 41–48 (2010)Google Scholar
- 7.Chuprat, S., Baruah, S.: Real-time divisible load theory: incorporating computation costs. In: Proceedings of the 17th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications, pp. 33–37 (2011)Google Scholar
- 8.Drozdowski, M.: Scheduling for Parallel Processing. Springer, London (2009)zbMATHCrossRefGoogle Scholar
- 9.Hu, M., Luo, J., Veeravalli, B.: Optimal provisioning for scheduling divisible loads with reserved cloud resources. In: Proceedings of the IEEE International Conference on Networks (ICON), pp. 204–209 (2012)Google Scholar
- 10.Jia, J., Veeravalli, B., Weissman, J.: Scheduling multisource divisible loads on arbitrary networks. IEEE Trans. on Parallel and Distrib. Systems 21, 520–530 (2010)CrossRefGoogle Scholar
- 11.Kumar, S., Dutta, K., Mookerjee, V.: Maximizing business value by optimal assignment of jobs to resources in grid computing. European J. of Oper. Res. 194, 856–872 (2009)MathSciNetzbMATHCrossRefGoogle Scholar
- 12.Lin, W., Liang, C., Wang, J.Z., Buyya, R.: Bandwidth-aware divisible task scheduling for cloud computing. Software: Practice and Experience (accepted, available online, 2013)Google Scholar
- 13.Sohn, J., Robertazzi, T.G., Luryi, S.: Optimizing computing costs using divisible load analysis. IEEE Trans. Parallel and Distrib. Systems 9, 225–234 (1998)CrossRefGoogle Scholar
- 14.Robertazzi, T.G.: Ten reasons to use divisible load theory. IEEE Computer 36, 63–68 (2003)CrossRefGoogle Scholar
- 15.van Hoesel, S., Wagelmans, A., Moerman, B.: Using geometric techniques to improve dynamic programming algorithms for the economic lot-sizing problem and extensions. European J. Oper. Res. 75, 312–331 (1994)zbMATHCrossRefGoogle Scholar
- 16.Yu, J., Buyya, R., Ramamohanarao, K.: Workflow Scheduling Algorithms for Grid Computing. In: Xhafa, F., Abraham, A. (eds.) Meta. for Sched. in Distri. Comp. Envi. SCI, vol. 146, pp. 173–214. Springer, Heidelberg (2008)Google Scholar