Abstract
A computer network provides the means for load sharing between processors. Most optimal load sharing (LS) algorithms for distributed systems require information about the sensitivity of the performance measure with respect to job flows. This information is generally difficult to obtain for real-time applications, due to the absence of closed-form expressions for performance as a function of flows. In this paper we present a method for estimating the mean response time gradients used in our LS-algorithm, based on a technique known as perturbation analysis (PA). Experimental results included demonstrate the adaptivity of the LS-algorithm to changing workload.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
K.M. Baumgartner, B.W. Wah, “A Global Load Balancing Strategy For A Distributed Computer System”, Proc. Workshop on the Future Trends of Distributed Computing Systems in the 1990s, Hong Kong, 1988.
T. Casavant, J. Kuhl, “A Taxonomy of Scheduling in General-Purpose Distributed Computing Systems”, IEEE Trans. Software Eng., vol. 14, pp. 141–154, Feb.1988.
X.R. Cao, “Convergence of parameter sensitivity estimates in a stochastic experiment”, IEEE Trans. Autom. Control, vol. 30, pp. 834–843, 1985.
X.R. Cao, “First order perturbation analysis of a simple multiclass finite source queue”, Performance Evaluation, vol. 7, pp. 31–41, 1987.
D.R. Cheriton, M. Stumm, “The Multi-Satellite Star: Structuring Parallel Computations for a Workstation Cluster”, Distributed Computing, 1988.
S. Ciereszko, “Design alternatives of adaptive load sharing servers in distributed systems”, Technischer Bericht, Institut für Betriebssysteme und Computerkommunikation, Technische Universität Dresden, 1992.
F. Douglis, J. Ousterhout, “Process Migration in the Sprite Operating System”, Proc. 7th Int. Conf. Distributed Computing Systems, Berlin, pp. 18–25, Sept.1987.
D.L. Eager, E.D. Lazowska, J. Zahoijan, “Adaptive Load Sharing in Homogeneous Distributed Systems”, IEEE Trans. Software Eng., vol. 12, pp. 662–675, May 1986.
R. Finkel, U. Manber, “DIB - A Distributed Implementation of Backtracking”, ACM Trans. Programming Languages and Systems 9(2), pp. 235–256, 1987.
C. Förster, “Adaptive Allocation of Computational Requirements to Heterogeneous Networks”, Kommunikation in Verteilten Systemen, N. Gerner, O. Spaniol (eds.), Springer- Verlag, pp. 324–337, Berlin, 1989.
R.G. Gallager, “A minimum delay routing algorithm using distributed computation”, IEEE Trans. Commun., vol. 25, pp. 73–85, Jan. 1977.
W.B. Gong, C.G. Cassandras, J. Pan, “Perturbation Analysis of a Multiclass Queueing System with Admission Control”, IEEE Trans. Autom. Control, vol. 36, no. 6, pp. 707–723, June 1991.
W.B. Gong, Y.C. Ho, “Smoothed (conditional) perturbation analysis of discrete event dynamic systems”, IEEE Trans. Autom. Control, vol. 32, no. 10, pp. 858–866, 1987.
P. Heidelberger, D. Towsley, “Sensitivity Analysis from Sample Paths using Likelihoods”, Management Science, vol. 35, no. 12, pp. 1475–1488, Dec. 1989.
Y.C. Ho, “Performance evaluation and perturbation analysis of discrete event dynamic systems”, IEEE Trans. Autom. Control, vol. 32, no. 7, pp. 563–572, 1987.
Y.C. Ho, S. Li, “Extensions of infinitesimal perturbation analysis”, IEEE Trans. Autom. Control, vol. 33, no. 5, pp. 427–438, 1988.
C. Kim, H. Kameda, “Optimal Static Load Balancing of Multi-class Jobs in a Distributed Computer System”, Proc. 10th Int. Conf. Distributed Computing Systems, Paris, 1990.
L. Kleinrock, “Queueing Systems Volume 1: Theory”, New York: Wiley, 1975.
L. Kleinrock, “On Distributed Systems Performance”, Computer Networks and ISDN Systems (20), pp. 209–215, 1990.
L. Kleinrock, W. Korfhage, “Collecting Unused Processing Capacity: An Analysis of Transient Distributed Systems”, Proc. 9th Int. Conf. Distributed Computing Systems, Newport Beach, California, June 1989.
K. Lee, D. Towsley, “An optimal threshold-based load balancing algorithm for quasi-static environments”, Tech. Report, Depart, of Comp, and Inform. Syst., Univ. of Massachusetts, 1987.
M. Livny, M. Melman, “Load Balancing in Homogeneous Broadcast Distributed Systems”, Proc. ACM Computer Network Performance Symposium, pp. 47–55, 1982.
M.W. Mutka, M. Livny, “The available capacity of a privately owned workstation environment”, Performance Evaluation (12), pp. 269–284, 1991.
D. Nichols, “Using Idle Workstations in a Shared Computing Environment”, Proc. 11th ACM Symp. on Operating System Principles, 1987.
S. Pulidas, D. Towsley, J.A. Stankovic, “Imbedding Gradient Estimators in Load Balancing Algorithms”, Proc. 8th Int. Conf. Distributed Computing Systems, pp. 482–490, 1988.
M.I. Reiman, A. Weiss, “Sensitivity analysis for simulation via likelihood ratios”, Proc. 1986 Winter Simulation Conf., pp. 285–289, 1986.
M.S. Squantale, R.D. Nelson, “Analysis of Task Migration in Shared-Memory Multiprocessor Scheduling”, Proc. 1991 ACM SIGMETRICS Conf. Measurement and Modeling of Computer Systems, San Diego, California, May 1991.
M. Stumm, “The design and implementation of a decentralized scheduling facility for a workstation cluster”, Proc. 2nd IEEE Conf. Computer Workstations, pp. 12–22, March 1988.
R. Suri, “Perturbation analysis: The state of the art and research issues explained via the G/G/l queue”, Proc. IEEE, vol. 77, no. 1, pp. 114–137, 1989.
M.M. Theimer, K.A. Lantz, “Finding Idle Machines in a Workstation-Based Distributed System”, IEEE Trans. Software Eng., vol.15, no.11, pp. 1444–1457, 1989.
Y.T. Wang, R.J.T. Monis, “Load Sharing in Distributed Systems”, IEEE Trans. Comput., vol. 34, pp. 204–217, Mar. 1985.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1993 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ciereszko, S., Hofmann, U. (1993). Adaptive load sharing with on-line gradient estimating in network environments. In: Gerner, N., Hegering, HG., Swoboda, J. (eds) Kommunikation in Verteilten Systemen. Informatik aktuell. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-78091-2_15
Download citation
DOI: https://doi.org/10.1007/978-3-642-78091-2_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-56482-9
Online ISBN: 978-3-642-78091-2
eBook Packages: Springer Book Archive