In the thesis, we propose a scheduling framework for online mixed-parallel workflows in heterogeneous multi-cluster environments, named Mixed-Parallel Online Workflow Scheduling (MOWS), which divides the entire scheduling process into four phases: task prioritizing, waiting queue scheduling, task rearrangement, and task allocation. Four new methods, shortest-workflow-first, priority-based backfilling, preemptive task execution and All-EFT task allocation, were developed for scheduling online mixed-parallel workflows under the MOWS framework.
The shortest-workflow-first strategy enforces the SJF policy [30] in the waiting queue scheduling phase in order to reduce the average makespan of all workflows.
The priority-based backfilling was introduced to allow out-of-order execution among tasks to improve resource utilization and thus the overall system performance. The preemptive task execution was developed for the task allocation phase to cooperate with SWS [9] used in the task prioritizing phase to take the advantages of both SWS [9] and CPWS [12]. The All-EFT for the task allocation phase always considers each cluster in the system and allocates the task to the cluster leading to the earliest estimated finish time.
We provide detailed examples for illustrating the superiority of the proposed methods over existing approaches. In addition, we conducted a series of simulation studies for performance evaluation and compared MOWS with a previously proposed approach in the literature called OWM. The experimental results indicate that each of the four proposed methods outperforms existing approaches significantly even under inaccurate estimation of task execution time. In average, MOWS can achieve around
70
16% performance improvement over OWM in terms of average makespan and SLR.
In the future, under the framework of MOWS there might be several research directions to further improve the scheduling performance of online mixed-parallel workflows in heterogeneous multi-cluster environments. For example, the preemptive task execution method could be extended to consider multiple running tasks for preemption simultaneously. This would increase the probability for high-priority tasks to start execution earlier and thus improve the overall system performance. For the shortest-workflow-first policy, other metrics for prioritizing workflows could be investigated in addition to the remaining execution time used in this thesis.
71
Reference
[1]. M. L. Pinedo, “Scheduling: Theory, Algorithms, and Systems”, Springer Publishing Company, 2008.
[2]. H. Topcuoglu, S. Hariri and M. Wu, “Performance-Effective and Low-Complexity Task Scheduling for Heterogeneous Computing”, IEEE Trans. on Parallel and Distributed Systems, vol. 13, pp. 260-274, 2002.
[3]. L.F. Bittencourt, R. Sakellariou and E.R.M. Madeira, “DAG Scheduling Using a Lookahead Variant of the Heterogeneous Earliest Finish Time Algorithm”, Proceedings of the 18th Euromicro Conference on Parallel, Distributed and Network-Based Processing, pp.27-34, 2010.
[4]. Y. Kwok and I. Ahmad. “Dynamic Critical-Path Scheduling: An Effective Technique for Allocation Task Graphs to Multi-processors”, IEEE Trans.
Parallel and Distributed Systems, vol. 7, pp. 506-521, 1996.
[5]. Z. Yu and W. Shi, “An Adaptive Rescheduling Strategy for Grid Workflow Applications”, Parallel and Distributed Processing Symposium, 2007. IEEE International, pp. 1-8, 2007.
[6]. A.M.A Ghanem, A.I. Saleh and H.A. Ali, “High Performance Adaptive Framework for Scheduling Grid Workflow Applications”, Computer Engineering and Systems (ICCES), International Conference on, pp. 52-57, 2010.
[7]. R. Sakellariou and H. Zhao, “A hybrid heuristic for DAG scheduling on heterogeneous systems”, In Proceedings of 13th Heterogeneous Computing Workshop (HCW 2004), vol. 2, pp.111b, 2004.
[8]. H. Zhao and R. Sakellariou, “Scheduling multiple DAGs onto heterogeneous systems”, Parallel and Distributed Processing Symposium, 2006. IPDPS 2006.
20th International, pp. 14, 2006.
72
[9]. Z. Yu and W. Shi, “A planner-Guided Scheduling Strategy for Multiple Workflow Applications”, On Parallel Processing Workshops, ICPP-W 08, pp. 8-12, 2008.
[10]. M. J. Quinn, “Parallel Programming in C with Mpi and Openmp”, McGraw-Hill Education, 2008.
[11]. K. C. Huang, “On Effects of Resource Fragmentation on Job Scheduling Performance in Computing Grids”, 2009 10th International Symposium on Pervasive Systems, Algorithms, and Networks, pp.701-705, 2009.
[12]. C.C. Hsu, K.C. Huang and F.J. Wang, “Online scheduling of workflow applications in grid environments”, Future Generation Computer Systems 27, pp.
860–870, 2011.
[13]. A.W. Mu‟alem and D.G. Feitelson, “Utilization, Predictability, Workloads, and User Runtime Estimates in Scheduling the IBM SP2 with Backfilling”, IEEE Trans. on Parallel and Distributed Systems, vol. 12, pp. 529-543, 2001.
[14]. R. M. Fujimoto, “Parallel discrete event simulation”, WSC '89 Proceedings of the 21st conference on Winter simulation, pp. 19-28, 1989.
[15]. U. Lublin and D. G. Feitelson, "The workload on parallel supercomputers:
modeling the characteristics of rigid jobs", Journal of Parallel and Distributed Computing, vol. 63, pp. 1105-1122, 2003.
[16]. M. Barreto, R. Avila and P. Navaux, “The MultiCluster Model to the Integrated Use of Multiple Workstation Clusters”, 3rd Workshop on Personal Computerbased Networks of Workstations, pp. 71–80, 2000.
[17]. M. Wieczorek, A. Hoheisel and R. Prodan, “Towards a general model of the multi-criteria workflow scheduling on the grid”, Future Generation Computer Systems 25, pp. 237-256, 2009.
[18]. J. Yu, R. Buyya and K. Ramamohanarao, “Workflow Scheduling Algorithms for Grid Computing”, Tech. Rep., GRIDS-TR-2007-10, University of Melbourne,
73
2010.
[19]. R. Sakellariou, H. Zhao, E. Tsiakkouri and M. Dikaiakos, “Scheduling Workflows with Budget Constraints”, In S.Gorlatch, M.Danelutto (Eds.), Integrated Research in Grid Computing, CoreGrid series, Springer-Verlag, to appear, 2005.
[20]. T. Tannenbaum, D. Wright, K. Miller and M. Livny, “Condor - A Distributed Job Scheduler”, Computing with Linux, The MIT Press, MA, 2002.
[21]. M . Wu and D. Gajski, “Hypertool: A Programming Aid for Message Passing Systems”, IEEE Trans. on Parallel and Distributed Systems, vol. 1, pp. 330-343, 1990.
[22]. T. Yang and A. Gerasoulis, “DSC: Scheduling Parallel Tasks on an Unbounded Number of Processors”, IEEE Trans. on Parallel and Distributed Systems, vol. 5, pp. 951-967, 1994.
[23]. S. Darbha and D.P. Agrawal, “Optimal Scheduling Algorithm for Distributed Memory Machines”, IEEE Trans. On Parallel and Distributed Systems, vol. 9, pp. 87-95, 1998.
[24]. G. Park, B. Shirazi and J. Marquis, “DFRN: A New Approach for Duplication Based Scheduling for Distributed Memory Multi-processor Systems”, Proc. Int'l Conf. Parallel Processing, pp. 157-166, 1997.
[25]. R. Bajaj and D.P. Agrawal, “Improving Scheduling of Tasks in a Heterogeneous Environment”, IEEE Trans. on Parallel and Distributed Systems, vol. 15, pp.
107-118, 2004.
[26]. M. Resende and C. Ribeiro, “Greedy Randomized Adaptive Search Procedures, State-of-the-art Handbook in MetaHeuristics”, Glover and Kochenberger, eds., Kluwer Academic Publishers, 2002.
[27]. H. Singh and A. Youssef, “Mapping and Scheduling Heterogeneous Task Graphs
74
Using Genetic Algorithms”, Proc. Heterogeneous Computing Workshop, pp.
86-97, 1996.
[28]. A. YarKhan and J. Dongarra, “Experiments with Scheduling Using Simulated Annealing in a Grid Environment”, Workshop on Grid Computing, pp.232-242, 2002.
[29]. J. Blythe, S. Jain, E. Deelman, Y. Gil, K. Vahi, A. Mandal and K. Kennedy, “Task Scheduling Strategies for Workflow-based Applications in Grid”, Cluster Computing and the Grid, IEEE International Symposium, vol. 2, pp. 759-767, 2005.
[30]. A. Silberschatz, J. Peterson, and P. Galvin, "Operating System Concepts", ed:
Addison-Wesley Publishing Company, pp. 155-156, 1991.
[31]. D.G. Feitelson and A.W. Mu‟alem, “Utilization and Predictability in Scheduling the IBM SP2 with Backfilling”, Proc. 12th Int'l Parallel Processing Symp., pp.
542-546, 1998.