Fast networks have made it possible to aggregate distributed CPU, memory, storage, and data to provide the potential for application performance superior to that attainable on any single system. However, achieving such performance on these metacomputing systems has proved to be difficult. Experience with the I-WAY [DFP + ss] and other metacomputing platforms demonstrates that effective application scheduling is critical to the achievement of performance for metacomputing applications. Currently, application developers develop customized application schedules to achieve performance on a metacomputer. Such application-centric schedules promote the performance of the application by evaluating system performance in terms of application resource requirements. To formalize and generalize the, as yet, ad hoc notion of application-centric scheduling emerging from the practices of metacomputing application developers [EMRP, SAR, GWP93], we are developing metacomputing scheduling agents calle...

Abstract not found

Grid computing is a promising technology for future computing platforms. Here, the task of scheduling computing resources proves difficult as resources are geographically distributed and owned by individuals with different access and cost policies. This paper addresses the idea of applying economic models to the scheduling task. To this end a scheduling infrastructure and a market-economic method is presented. The efficiency of this approach in terms of response- and wait-time minimization as well as utilization is evaluated by simulations with real workload traces. The evaluations show that the presented economic scheduling algorithm provides similar or even better average weighted response-times as common algorithms like backfilling. This is especially promising as the presented economic models have additional advantages as e.g. support for different price models, optimization objectives, access policies or quality of service demands.

Due to the increasing sophistication and miniaturization of computational components, complex, distributed systems of interacting agents are becoming ubiquitous. Such systems, where each agent aims to optimize its own performance, but where there is a welldefined set of system-level performance criteria, are called collectives. The fundamental problem in analyzing/designing such systems is in determining how the combined actions of a large number of agents leads to “coordinated ” behavior on the global scale. Examples of artificial systems which exhibit such behavior include packet routing across a data network, control of an array of communication satellites, coordination of multiple rovers, and dynamic job scheduling across a distributed computer grid. Examples of natural systems include ecosystems, economies, and the organelles within a living cell. No current scientific discipline provides a thorough understanding of the relation between the structure of collectives and how well they meet their overall performance criteria. Although still very young, research on collectives has resulted in successes both in understanding and designing such systems. It is expected that as it matures and draws upon other disciplines related to collectives, this field will greatly expand the range of computationally addressable tasks. Moreover, in addition to drawing on them, such a fully developed field of collective intelligence may provide insight into already established scientific fields, such as mechanism design, economics, game theory, and population biology. This chapter provides a survey to the emerging science of collectives.

We believe that large-scale replica management solutions should be based on an economic model. In this paper, we discuss the benefits provided by an economic approach and outline important directions for future research.

Market mechanisms evolved in a human social context where they are part of a highly effective and robust practice of distributed decision making concerning the allocation of resources and coordination of economic activity. Recently there has been more interest in applying market organization to computational systems. This paper defines and discusses some concepts related to the operation of markets, focusing on the role of mechanisms. A distinction is introduced between natural and programmed markets. The differences are discussed, with examples drawn from recent computer science literature. Keywords: Market Mechanism, Programmed Market, Resource Allocation, Equilibration 1. INTRODUCTION Markets have long been a feature of human culture. A farmer's produce market, a used car lot, an art auction house and a stock exchange are all familiar examples of market institutions. As information technology improves our ability to communicate rapidly and easily across great distances and make s...

Grid computing can be defined as coordinated resource sharing and problem solving in dynamic, multiinstitutional collaborations [1]. As more Grids are deployed worldwide, the number of multi-institutional collaborations is rapidly growing. However, for Grid computing to realize its full potential, it is expected that Grid participants are able to use one another resources. Resource negotiation (i.e. exchange or trading of resources between Grids) enables Grid participants to face an unstable request environment. The aim of this position paper is to present a survey of the current state and challenges of resource negotiation research, with a Machine Learning perspective. We support the view that negotiation and learning are intrinsically linked. In particular, we show the expected benefits of integrating Machine Learning techniques with resource negotiation.