A Distributed and Parallel Processing Laboratory

点击翻译按钮获取中文摘要

Through this project, an instructional laboratory is being established to provide undergraduate computer science students with hands-on experience in distributed and parallel processing techniques. While most undergraduate students study various aspects of distributed and parallel computing in various courses during their undergraduate experience, very few actually receive hands-on experience on such systems. Most general-purpose computing laboratories are not designed for the type of system configuration and re-configuration needed to illustrate the various concepts involved in distributed and parallel computing. In addition, the generally low availability and high cost of such systems usually precludes hands-on experimentation. The result of this situation is that students graduate and proceed on to professional practice with a strong tendency to favor the use of centralized computing techniques over distributed computing techniques. One side-effect of this situation is that distributed and parallel systems, despite their high potentials, are largely underutilized in today's computing technology. The need for computer professionals with distributed computing expertise can only increase as distributed processing systems such as the Internet and the World Wide Web continue to make rapid gains in utilization. The system established in this project provides, at reasonable cost, a system that can be used to provide hands-on exposure to distributed and parallel computing concepts in three primary undergraduate courses: Parallel Processing, Operating Systems, and Data Structures and Algorithms. A number of other undergraduate courses may potentially benefit from the laboratory. The system is organized around one Sun SPARC station_20 with four processors and four uniprocessor Sun SPARC station_4 workstations. The choice of Sun workstations is consistent with current industrial/academic practice.