
This Article  
 
Share  
Bibliographic References  
Add to:  
Digg Furl Spurl Blink Simpy Del.icio.us Y!MyWeb  
Search  
 
ASCII Text  x  
ChaoJu Hou, K.G. Shin, "Load Sharing with Consideration of Future Task Arrivals in Heterogeneous Distributed RealTime Systems," IEEE Transactions on Computers, vol. 43, no. 9, pp. 10761090, September, 1994.  
BibTex  x  
@article{ 10.1109/12.312127, author = {ChaoJu Hou and K.G. Shin}, title = {Load Sharing with Consideration of Future Task Arrivals in Heterogeneous Distributed RealTime Systems}, journal ={IEEE Transactions on Computers}, volume = {43}, number = {9}, issn = {00189340}, year = {1994}, pages = {10761090}, doi = {http://doi.ieeecomputersociety.org/10.1109/12.312127}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, }  
RefWorks Procite/RefMan/Endnote  x  
TY  JOUR JO  IEEE Transactions on Computers TI  Load Sharing with Consideration of Future Task Arrivals in Heterogeneous Distributed RealTime Systems IS  9 SN  00189340 SP1076 EP1090 EPD  10761090 A1  ChaoJu Hou, A1  K.G. Shin, PY  1994 KW  resource allocation; distributed processing; probability; queueing theory; realtime systems; Bayes methods; parameter estimation; state estimation; minimisation; load sharing algorithm; future task arrivals; heterogeneous distributed realtime systems; unguaranteed task transfer; abundant resources; best receiving node; minimumlaxityfirstserved discipline; remote node location policy; Bayesian analysis; probability; queueing analysis; online Bayesian parameter estimation; timestamped regionchange broadcasts; dynamically varying workloads; computational overhead; simulation; dynamic failure; task collisions; excessive task transfers; deadlines; performance evaluation. VL  43 JA  IEEE Transactions on Computers ER   
In a heterogeneous distributed realtime system, transferring an unguaranteed task at a node to another node currently with the most abundant resources is not necessarily the best decision. We propose a new load sharing (LS) algorithm for realtime applications which takes into account the effect of future task arrivals on locating the best receiver for each unguaranteed task. Upon arrival of a task at a node, the node first checks whether it can complete the task in time using the minimumlaxityfirstserved discipline. If the node cannot guarantee the arrived task, or if some of existing guarantees were to be invalidated as a result of inserting the task into its queue, then the node must locate a remote node to which each unguaranteed task is to be transferred. The LS algorithm minimizes not only the probability of transferring an unguaranteed task T to an incapable node with Bayesian analysis, but also the probability that a remote node fails to guarantee T because of future arrivals of tighterlaxity tasks with queueing analysis. All parameters needed for a node's LS decision are collected/estimated online using timestamped regionchange broadcasts (TSRCBs) and Bayesian estimation. By using TSRCBs, the collected state information can be used to estimate other nodes' states. Use of Bayesian estimation makes the LS algorithm adaptive to dynamically varying workloads with little computational overhead. Simulation results show that the proposed LS algorithm outperforms other LS algorithms in minimizing the probability of dynamic failure, task collisions and excessive task transfers.
[1] K. G. Shin and C.J. Hou, "Design and evaluation of effective load sharing in distributed realtime systems,"IEEE Trans. Parallel Distrib. Syst., vol. 5, no. 7, pp. 704719, July 1994.
[2] C. M. Krishna and K. G. Shin, "Performance measures for multiprocessor controllers," inPerformance '83, A. K. Agrawala and S. K. Tripathi, Eds. New York: NorthHolland, 1983, pp. 229250.
[3] K.G. Shin, C.M. Krishna, and Y.H. Lee, "A Unified Method for Evaluating RealTime Computer Controllers and Its Application,"IEEE Trans. Automatic Control, Vol. AC30, No. 4, Apr. 1985, pp. 357366.
[4] J. Hong, X. Tan, and D. Towsley, "A performance analysis of minimum laxity and earliest deadline scheduling in a realtime system,"IEEE Trans. Comput., vol. 38, no. 12, pp. 17361744, Dec. 1989.
[5] D. Eager, E. Lazowska, and J. Zahorjan, "Adaptive load sharing in homogeneous distributed systems,"IEEE Trans. Software Eng., vol. SE12, no. 5, pp. 662675, May 1986.
[6] J. A. Stankovic, "Simulations of three adaptive, decentralized controlled, job scheduling algorithms,"Comput. Networks, pp. 199217, June 1984.
[7] J. A. Stankovic, "An application of Bayesian decision theory to decentralized control of job scheduling,"IEEE Trans. Comput., vol. C34, no. 2, pp. 117130, Feb. 1985.
[8] K. G. Shin and Y.C. Chang, "Load sharing in distributed realtime systems with state change broadcasts,"IEEE Trans. Comput., vol. 38, no. 8, pp. 11241142, Aug. 1989.
[9] K. G. Shin and Y.C. Chang, "Load sharing in hypercube multicomputers for realtime applications," presented at the4th Conf. Hypercube, Concurrent Comput., and Applicat., 1989.
[10] J. F. Kurose and R. Chipalkatti, "Load sharing in soft realtime distributed computer systems,"IEEE Trans. Comput., vol. C36, pp. 9931000, Aug. 1987.
[11] J. A. Stankovic, K. Ramamritham, and S. Chang, "Evaluation of a flexible task, scheduling algorithm for distributed hard realsystems,"IEEE Trans. Comput., vol. C34, no. 12, pp. 11301141, Dec. 1985.
[12] K. Ramamritham, J. A. Stankovic, and W. Zhao, "Distributed scheduling of tasks with deadlines and resource requirements,"IEEE Trans. Comput., vol. 38, no. 8, pp. 11101141, Aug. 1989.
[13] T. P. Yum and M. Schwartz, "The joinbiasedqueue rule and its application to routing in computer communication networks,"IEEE Trans. Commun., vol. COM29, no. 4, pp. 505511, Apr. 1981.
[14] T. P. Yum and H.C. Lin, "Adaptive load balancing for parallel queues with traffic constraints,"IEEE Trans. Commun., vol. COM32, no. 12, pp. 13391342, Dec. 1984.
[15] R. Mirchandaney, D. Towsley, and J. A. Stankovic, "Adaptive load sharing in heterogeneous systems," inProc. 9th Int. Conf. Distrib. Computing Syst., pp. 298306, 1989.
[16] S. Pulidas, D. Towsley, and J. A. Stankovic, "Imbedding gradient estimators in load balancing algorithms," inProc. 8th Int. Conf. Distributed Comput. Syst., 1988, pp. 482490.
[17] L. Kleinrock,Queueing Systems, Volume I: Theory. New York: John Wiley, 1975.
[18] D. Gross and C. Harris,Fundamentals of Queueing Theory, second ed. New York: John Wiley, 1985.
[19] S. M. Ross,Stochastic Processes. New York: John Wiley, 1983.
[20] M. H. DeGroot, "Optimal Statistical Decisions. New York: McgrawHill, 1970.
[21] R. Mirchandaney, D. Towsley, and J. A. Stankovic, "Analysis of the effect of delays on load sharing,"IEEE Trans. Comput., vol. 38, no. 11, pp. 15131525, Nov. 1989.
[22] M. H. DeGroot, "Probability and Statistics, second ed. Reading, MA: AddisonWesley, 1986.