Abstract
A scheduling algorithm is proposed for large-scale, heterogeneous distributed systems working on SPMD tasks with homogeneous input. The new algorithm is based on stochastic optimization using a modified least squares method for the identification of communication and performance parameters. The model of computation involves a server distributing tasks to clients. The goal of the optimization is to reduce execution time by the clients. The costs of getting the task from the server, execution of the task and sending the results back are estimated; and the scheduling is based on adaptive division of work (input for the clients) into blocks.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Granichin, O., Polyak, B.: Randomized algorithms of estimation and optimization under almost arbitrary noise. M. Nauka (2003)
Nakano, A.: High performance computing and simulations (Spring ’07). Available online: http://cacs.usc.edu/education/cs653.html
Weissman, J.: Prophet: automated scheduling of SPMD programs in workstation networks. Concurrency: Practice and Experience 11(6), 301–321 (1999)
Cermele, M., Colajanni, M., Necci, G.: Dynamic load balancing of distributed SPMD computations with explicit message-passing. In: Proc. of the IEEE Workshop on Heterogeneous Computing, pp. 2–16 (1997)
He, Y., Hsu, W., Leiserson, C.: Provably efficient adaptive scheduling for parallel jobs. In: The Proc. of the 12th Workshop on Job Scheduling Strategies for Parallel Processing (2006)
Ichikawa, S., Yamashita, S.: Static load balancing of parallel PDE solver for distributed computing environment. In: Proc. ISCA 13th Int’l. Conf. Parallel and Distributed Computing Systems (PDCS-2000), pp. 399–405 (2000)
Casavant, T., Kuhl, J.: A taxonomy of scheduling in general-purpose distributed computing systems. IEEE Trans. on Software Engineering 14(2), 141–154 (1988)
Blumofe, R., Leiserson, C.: Scheduling multithreaded computations by work stealing. In: Proc. 35th Annual IEEE Conf. on Foundations of Computer Science (FOCS’94), Santa Fe, New Mexico. IEEE Computer Society Press, Los Alamitos (1994)
Neill, D., Wierman, A.: On the benefits of work stealing in shared memory multiprocessors. report, http://www.cs.cmu.edu/acw/15740/paper.pdf
Anderson, C., et al.: SETI@Home: an experiment in public-resource computing. Comm. of the ACM 45(11), 56–61 (2002)
Jaillet, C., Krajecki, M.: Constructing optimal Golomb rulers in parallel. In: Proc. 6th European Workshop on OpenMP, pp. 29–34 (2004)
Warren, M., Salmon, J.: A parallel hashed oct-tree n-body algorithm. Supercomputing, 12–21 (1993)
Lof, H.: Iterative and adaptive PDE solvers for shared memory architectures. Acta universitalis: digital comprehensive summaries of Uppsala dissertations from the faculty of science and technology (2006)
Hamidzadeh, B., Lilja, D.: Dynamic scheduling strategies for shared memory multiprocessors. In: International Conference on Distributed Computing Systems, pp. 208–215 (1996)
Autonomic computing: IBM’s perspective on the state of information technology, http://www.research.ibm.com/autonomic/manifesto/autonomic_computing.pdf
Kushner, H., Yin, G.: Stochastic approximation and recursive algorithms and applications, 2nd edn. Springer, Heidelberg (2003)
Estkover, C.A. (ed.): Chaos and Fractals: A Computer Graphical Journey, p. 468. Elsevier Science, Amsterdam (1998)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Panshenskov, M., Vakhitov, A. (2007). Adaptive Scheduling of Parallel Computations for SPMD Tasks. In: Gervasi, O., Gavrilova, M.L. (eds) Computational Science and Its Applications – ICCSA 2007. ICCSA 2007. Lecture Notes in Computer Science, vol 4706. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-74477-1_4
Download citation
DOI: https://doi.org/10.1007/978-3-540-74477-1_4
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-74475-7
Online ISBN: 978-3-540-74477-1
eBook Packages: Computer ScienceComputer Science (R0)