Queueing dynamics and maximal throughput scheduling in switched processing systems

被引:57
作者
Armony, M [1 ]
Bambos, N
机构
[1] NYU, Stern Sch Business, Dept Informat Operat & Management Sci, New York, NY 10012 USA
[2] Stanford Univ, Dept Management Sci & Engn, Stanford, CA 94305 USA
[3] Stanford Univ, Dept Elect Engn, Stanford, CA 94305 USA
基金
美国国家科学基金会;
关键词
switched processing systems; dynamic scheduling; throughput maximization; cone policies; adaptive batching policies; trace-based modeling;
D O I
10.1023/A:1024714024248
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
We study a processing system comprised of parallel queues, whose individual service rates are specified by a global service mode ( configuration). The issue is how to switch the system between various possible service modes, so as to maximize its throughput and maintain stability under the most workload-intensive input traffic traces ( arrival processes). Stability preserves the job inflow - outflow balance at each queue on the traffic traces. Two key families of service policies are shown to maximize throughput, under the mild condition that traffic traces have long-term average workload rates. In the first family of cone policies, the service mode is chosen based on the system backlog state belonging to a corresponding cone. Two distinct policy classes of that nature are investigated, MaxProduct and FastEmpty. In the second family of batch policies ( BatchAdapt), jobs are collectively scheduled over adaptively chosen horizons, according to an asymptotically optimal, robust schedule. The issues of nonpreemptive job processing and non-negligible switching times between service modes are addressed. The analysis is extended to cover feed-forward networks of such processing systems/nodes. The approach taken unifies and generalizes prior studies, by developing a general trace-based modeling framework (sample-path approach) for addressing the queueing stability problem. It treats the queueing structure as a deterministic dynamical system and analyzes directly its evolution trajectories. It does not require any probabilistic superstructure, which is typically used in previous approaches. Probability can be superposed later to address finer performance questions (e.g., delay). The throughput maximization problem is seen to be primarily of structural nature. The developed methodology appears to have broader applicability to other queueing systems.
引用
收藏
页码:209 / 252
页数:44
相关论文
共 29 条
[1]  
ARMONY M, 1999, P 37 ANN ALL C COMM, P42
[2]  
ARMONY M, 1999, THESIS STANFORD U
[3]  
Baccelli F., 1994, Elements of Queueing Theory
[4]  
Bambos N., 1994, Queueing Systems Theory and Applications, V15, P137, DOI 10.1007/BF01189235
[5]   SCHEDULING AND STABILITY ASPECTS OF A GENERAL-CLASS OF PARALLEL PROCESSING SYSTEMS [J].
BAMBOS, N ;
WALRAND, J .
ADVANCES IN APPLIED PROBABILITY, 1993, 25 (01) :176-202
[6]  
Bell SL, 2001, ANN APPL PROBAB, V11, P608
[7]  
BOHM V, 1975, SIAM J APPL MATH, V28, P303, DOI 10.1137/0128026
[8]   Simple necessary and sufficient conditions for the stability of constrained processes [J].
Budhiraja, A ;
Dupuis, P .
SIAM JOURNAL ON APPLIED MATHEMATICS, 1999, 59 (05) :1686-1700
[9]   DISCRETE FLOW NETWORKS - BOTTLENECK ANALYSIS AND FLUID APPROXIMATIONS [J].
CHEN, H ;
MANDELBAUM, A .
MATHEMATICS OF OPERATIONS RESEARCH, 1991, 16 (02) :408-446
[10]  
Dai J. G., 2000, Proceedings IEEE INFOCOM 2000. Conference on Computer Communications. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Cat. No.00CH37064), P556, DOI 10.1109/INFCOM.2000.832229