Abstract
Increasingly, feedback of measured run-time information is being used in the optimization of computation execution. Because of this, the need for a model relating the static view of a computation to its runtime variance is becoming more important. Recently, we have described such a model which uses the notion of uncertainty to provide bounds on key scheduling parameters of the run-time computation. In this paper, we demonstrate how our model provides a foundation for robust parallel scheduling, i.e., scheduling that optimizes for computation execution in the presence of run-time variance. While this work was inspired by our previous study of uncertainty due to measurement intrusion, the scheduling paradigm presented here represents a broader, more general application of the uncertainty concept.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
M.S. Andersland and T.L. Casavant, “Recovering uncorrupted event traces from corrupted event traces in parallel/distributed systems,” 20th Int. Conf. on Par. Proc., St. Charles, IL, August 1991, pp. II:108–116.
T. Casavant and J. Kuhl, “A Taxonomy of Scheduling in General Purpose Distributed Computing Systems,” IEEE Trans. on Software Eng., Vol. 14, No. 2, February 1988, pp.141–154.
P.P. Chang, S.A. Mahlke, W.Y. Chen, N.J. Warter, and W.W. Hwu, “IMPACT: An Architectural Framework for Multiple-Instruction Issue Processors,” in Proc. of the 18th Annual Int. Symp. on Comp. Arch., May 1991, pp.266–275.
R.D. Dietz, T.L. Casavant, T.A. Braun, T.E. Scheetz, and M.S. Andersland, “Robust Scheduling of Parallel Computation via Static Enumeration and Dynamic Instantiation & Activation (SEDIA),” Technical Report, The University of Iowa, September 1996.
R.D. Dietz, T.L. Casavant, T.E. Scheetz, T.A. Braun, and M.S. Andersland, “Modeling the Impact of Run-Time Uncertainty on Optimal Computation Scheduling Using Feedback” in Proc. of the 26th Annual Int. Conference on Parallel Processing (to appear), August 1997.
H. El-Rewini, T.G. Lewis, and H.H. Ali, Task Scheduling in Parallel and Distributed Systems, Prentice-Hall, Inc., New Jersey, 0-13-099235-6, 1994.
A. Gerasoulis, J. Jiao, and T. Yang, “Experience with Graph Scheduling for Mapping Irregular Scientific Computation,” in Proc. of the First IPPS Workshop on Solving Irregular Problems on Dist. Memory Machines, Santa Barbara, CA, April 1995, pp. 1–8.
D. Heckerman, “A Bayesian Approach to Learning Causal Networks,” Technical Report MSR-TR-95-04, Microsoft Research, March, 1995.
R.V. Hogg and J. Ledolter, Engineering Statistics, Macmillan Publishing Company, New York, 1987.
U. Holzle, “Adaptive Optimization for SELF: Reconciling High Performance with Exploratory Programming,” Ph.D. Dissertation, Aug 1994.
B. Kruatrachue and T. Lewis, “Grain Size Determination For Parallel Processing,” IEEE Software, January 1988, pp.23–31.
L. Lamport, “Time, Clocks, and the Ordering of Events in a Distributed System,” Comm. of the ACM, Vol. 21, July 1978, pp.558–564.
Y. Ma, “Inductive Classifier Learning from Data: An Extended Bayesian Belief Function Approach,” Ph.D. Dissertation, University of Illinois, 1995.
C. McCreary and H. Gill, “Automatic Determination of Grain Size for Efficient Parallel Processing,” Comm. of the ACM, Vol. 32, No. 9, September 1989, pp.1073–1078.
S. Pande and K. Psarris, “Program Repartitioning on Varying Communication Cost Parallel Architectures,” Journal of Par. and Dist. Comp., March 1996, pp.205–213.
J. Pearl, Probabalistic Reasoning in Intelligent Systems, Morgan Kaufmann Publishers, Inc., San Francisco, California, 1988.
K. Pettis and R.C. Hansen, “Profile Guided Code Positioning,” in Proc. of the ACM SIGPLAN '90 Conf. on Prog. Language Design and Implementation, White Plains, New York, June 20–22, 1990, pp.16–27.
A.D. Samples, “Profile-Driven Compilation,” Ph.D. Dissertation, University of Berkeley, April 1991.
T.E. Scheetz, T.A. Braun, T.L. Casavant, J.A. Gannon, M.S. Andersland, and R.D. Dietz, “Effectiveness of Software Trace Recovery Techniques for Current Parallel Architectures,” Int. Conf. on High Performance Computing, New Delhi, India, December 27–30, 1995.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Dietz, R.D., Casavant, T.L., Scheetz, T.E., Braun, T.A., Andersland, M.S. (1997). Using run-time uncertainty to robustly schedule parallel computation. In: Malyshkin, V. (eds) Parallel Computing Technologies. PaCT 1997. Lecture Notes in Computer Science, vol 1277. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-63371-5_3
Download citation
DOI: https://doi.org/10.1007/3-540-63371-5_3
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63371-6
Online ISBN: 978-3-540-69525-7
eBook Packages: Springer Book Archive