Classification of load distribution algorithms

The allocation of workload in distributed systems has almost as many views as one would like. The attempt to find suitable structures to classify the different ways of solving this task can therefore reflect the underlying model only. In the following paper a three models is developed. The load model is used to describe the workload and its current distribution within the system. The action model defines for a given algorithm at any instant of time the eligible next step(s). And finally, the solution model shows the formal context for obtaining and evaluating the load distribution decisions. All three models and their interrelationship are illustrated by selected examples of existing load distribution approaches. The result of the following investigation is the recommendation that load distribution algorithms can be classified according the five criteria: objectives, type and amount of used information, the source of the distribution, the parameter time, and the initiating instance.

[1]  Thomas L. Casavant,et al.  A Taxonomy of Scheduling in General-Purpose Distributed Computing Systems , 1988, IEEE Trans. Software Eng..

[2]  Nicholas Carriero,et al.  Linda and Friends , 1986, Computer.

[3]  Erol Gelenbe,et al.  Dynamic load balancing in distributed systems , 1994, Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems.

[4]  D. K. Subramanian,et al.  An Affinity-Based Dynamic Load Balancing Protocol for Distributed Tansaction Processing Systems , 1993, Perform. Evaluation.

[5]  John A. Stankovic,et al.  An Application of Bayesian Decision Theory to Decentralized Control of Job Scheduling , 1985, IEEE Transactions on Computers.

[6]  Hisao Kameda,et al.  An algorithm for optimal static load balancing in distributed computer systems , 1992 .

[7]  Leslie Lamport,et al.  Time, clocks, and the ordering of events in a distributed system , 1978, CACM.

[8]  Mario Gerla,et al.  The advantage of dynamic tuning in distributed asymmetric systems , 1990, Proceedings. IEEE INFOCOM '90: Ninth Annual Joint Conference of the IEEE Computer and Communications Societies@m_The Multiple Facets of Integration.

[9]  Thomas Kunz,et al.  The Influence of Different Workload Descriptions on a Heuristic Load Balancing Scheme , 1991, IEEE Trans. Software Eng..

[10]  Philip S. Yu,et al.  A Performance Study of Robust Distributed Load Sharing Strategies , 1994, IEEE Trans. Parallel Distributed Syst..

[11]  Kentaro Shimizu,et al.  Adaptive bidding load balancing algorithms in heterogeneous distributed systems , 1994, Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems.

[12]  Giuseppe Serazzi,et al.  Workload characterization: a survey , 1993, Proc. IEEE.

[13]  Jeff Kramer,et al.  Methodical Analysis of Adaptive Load Sharing Algorithms , 1992, IEEE Trans. Parallel Distributed Syst..

[14]  S. Zhou,et al.  A Trace-Driven Simulation Study of Dynamic Load Balancing , 1987, IEEE Trans. Software Eng..