ITR: An Experimental Study of Thin-Client Computing Architectures

瘦客户机系统的日益普及使得开发分析和比较其性能的技术，评估瘦客户机计算模型在现代网络环境中的普遍可行性，以及确定控制这些体系结构性能的因素变得非常重要。这个项目涉及到进行一系列的实验，以获得信息和理解的有效性瘦客户端架构机制的图形和多媒体应用程序在各种网络环境。 这些实验包括使用几个真实的应用程序进行的性能分析，以及使用专门的微基准测试进行的受控实验，这些测试旨在隔离瘦客户机架构的特定方面，这些方面对整体应用程序性能有显著影响。应用程序和用户工作负载将有所不同，以了解和分析不同的远程显示机制在网络和服务器资源方面的可伸缩性。 这些实验将在一个测试环境中进行，其中网络特性，如带宽，延迟和数据包丢失以受控的方式变化，以了解它们对瘦客户端架构设计选择的影响。 这些实验的结果将有fundamentalscientific的影响，奠定了定量基础，为未来的创新，在瘦客户机计算的研究。

The recent and growing popularity of thin-client systems makes itimportant to develop techniques for analyzing and comparing theirperformance, to assess the general feasibility of the thin-clientcomputing model in modern networking environments, and to determinethe factors that govern the performance of these architectures. This project involves carrying out a series of experiments to gaininformation and understanding on the effectiveness of thin-clientarchitectural mechanisms for graphical and multimedia applications invarious network environments. These experiments include performanceanalysis using several real applications as well as controlledexperiments with specialized micro-benchmarks designed to isolateparticular aspects of thin-client architectures that have significantimpact on overall application performance. Application and userworkloads will be varied to understand and analyze the scalability ofdifferent remote display mechanisms in terms of network and serverresources. These experiments will be conducted in a testbedenvironment in which network characteristics such as bandwidth,latency, and packet loss are varied in a controlled manner tounderstand their impact on thin-client architectural designchoices. The results of these experiments will have a fundamentalscientific impact by laying a quantitative foundation for futureinnovations in thin-client computing research.