Abstract. Distributed computations, dealing with large amounts of data, are scheduled in Grid clusters today using either a task-centric mechanism, or a worker-centric mechanism. Because of the large data sets, the execution time is bounded by the cost of data transfer. In this paper, we introduce new worker-centric scheduling strategies that are novel in that they aim to implicitly exploit the locality of interest in order to reduce the cost of data transfer. Many Grid applications are characterized by such a locality of interest, i.e., a file is often accessed by multiple tasks and, more importantly, a set of files that are accessed by one task are also likely to be accessed together by other tasks. Our new deterministic, as well as probabilistic, scheduling algorithms implicitly exploit this feature to improve running time. Our experiments are done with traces of a real Grid application (Coadd), and show that our algorithms are able to achieve utilization of over 90%, while reducing makespan significantly compared to task-centric approaches.

Over the last decade, clusters have become the fastest growing platforms in high-performance computing. More recently, Grids were emerging as next generation computing platforms for large-scale computation and data intensive problems in industry, academic, and government organizations. Meanwhile, an increasing number of real-time applications running on clusters and Grids have mandatory security requirements in addition to stringent timing constraints. Conventional real-time scheduling algorithms developed for clusters and Grids, however, either disregard applications ’ security needs, and thus expose the applications to security threats, or run applications at inferior security levels without optimizing security performance. In recognition that many applications running on clusters and Grids demand both real-time performance and security, in this dissertation research we investigate the problem of scheduling real-time applications with various security requirements. First, we propose a security middleware model (or SMW for short) from which security-sensitive real-time applications are enabled to exploit a variety of security

iv ing between multiple parties encompasses and expands on much of the thinking that underlies agent-mediated e-commerce and on-demand computing systems. The role of the infrastructure in an open setting – as it applies to resource allocation mechanisms – is to ensure or verify the property of strategyproofness, namely, whether a self-interested agent can maximize her utility by simply reporting information about her preferences for different resource allocation truthfully. I present two approaches, with the role of the infrastructure slightly different in each. The first approach considers passive verification of the strategyproofness of mechanisms, while the second approach considers active enforcement of strategyproofness in decentralized auctions for dynamic resource allocation. I present monotonic resource estimation and pricing algorithms that can be used to ensure strategyproofness of a mechanism and empirical results from simulations using data collected from the Crimson Grid. Contents

In a large-scale distributed infrastructure, users and administrators typically desire to perform on-demand operations that act upon the most up-to-date state of the infrastructure. These on-demand operations range from monitoring the up-to-date machine properties in the infrastructure, to making Grid scheduling decisions for different tasks based on the current status of the infrastructure. However, the scale and dynamism present in the operating environment make it challenging to support these operations efficiently. This paper discusses several on-demand operations that we have been studying, challenges associated with them, and how to meet the challenges. Specifically, we build techniques for 1) on-demand group monitoring that allows users and administrators of an infrastructure to query and aggregate the up-to-date state of the machines (e.g., CPU utilization) in a group or multiple groups, 2) an on-demand Grid scheduling strategy that makes scheduling decisions based on the current availability of compute nodes, 3) an on-demand Grid computation strategy that chooses the best algorithm for the current input data set among multiple algorithms available. We also present our ongoing work. 1.