II-EN: Infrastructure to Support Desktop Virtualization Experiments for Research and Education

该项目是一个计算研究基础设施增强基础设施(CRI II-EN)项目，专注于测量和评估瘦客户端和虚拟桌面云？俄亥俄州立大学(VDC)。通过过渡他们的传统桌面？将专用硬件和软件安装到可通过瘦客户端访问的VDC，用户社区在用户便利性和成本节约方面受益匪浅。然而，要以可扩展且经济高效的方式分配和管理VDC资源，云服务提供商面临着独特的挑战。VDC中的用户工作负载配置文件是突发的(例如，在日常桌面启动期间，当用户在文本和图形密集型应用程序之间切换时)，并且瘦客户端用户体验质量(QOE)对互联网中的网络健康变化高度敏感。遗憾的是，现有的工作主要集中在基于CPU和内存负载的效用函数来管理服务器端资源，而没有考虑网络健康和瘦客户端用户体验。在VDC平台中，如果没有针对系统负载、网络运行状况和瘦客户端用户体验的综合实用程序信息进行资源分配，即使只有几十个用户，也不可避免地会导致代价高昂的资源过度配置。此外，由于缺乏从VDC的服务器端衡量用户体验的工具，VDC中的管理功能(如配置瘦客户端协议参数)通常是通过猜测来执行的，这反过来又会影响用户的QOE。为了解决在开发具有令人满意的瘦客户端用户QOE的可扩展VDC方面的上述研究挑战，PI开发了？VMLab？试点基础设施，可支持面向研究和教育社区的桌面虚拟化实验。在过去两年中，该基础设施支持：(A)与VDC资源分配和瘦客户端性能基准相关的研发活动；(B)面向系统管理员的桌面虚拟化沙箱；(C)面向教师和学生的课堂实验室的虚拟桌面；以及(D)评估在虚拟桌面中部署计算密集型交互应用程序(例如远程卷可视化)的可行性。该项目将扩展VMLab基础设施，并在生产性使用期间实现系统测量和用户体验评估。更广泛的影响该项目的更广泛的影响包括为高中生和来自代表性不足群体的其他用户提供服务。这项拟议的研究将探索使用瘦客户机计算模式为无法购买或维护完整PC的群体提供高质量用户体验的潜力。瘦客户端模型不仅在购买成本方面具有成本效益(假设最终用户不必摊销后端基础设施的成本，在此场景中就是这种情况)，而且由于设备维护和故障排除的简单性，还具有较低的稳定成本。托管在云中的桌面应用程序领域是一个重要且迅速崛起的领域(无论是在学术界还是在工业界)。成果的广泛传播将对一项具有经济价值的技术产生积极影响。

This project is a Computing Research Infrastructure Enhancing Infrastructure (CRI II-EN) project focused on measurement and evaluation of thin clients and ?virtual desktop clouds? (VDCs) at Ohio State University . By transitioning their ?traditional desktops? that have dedicated hardware and software installations to VDCs that are accessible via thin-clients, user communities greatly benefit in terms of user convenience, and cost-savings. However, to allocate and manage VDC resources in a scalable and cost-effective manner, cloud service providers are faced with unique challenges. User workload profiles in VDCs are bursty (e.g., during daily desktop startup, when user switches between text and graphic intensive applications), and thin-client user Quality of Experience (QoE) is highly sensitive to network health variations in the Internet. Unfortunately, existing works focus mainly on managing server-side resources based on utility functions of CPU and memory loads, and do not consider network health and thin-client user experience. Resource allocations without combined utility-directed information of system loads, network health and thin-client user experience in VDC platforms inevitably results in costly over-provisioning of resources, even for as few as tens of users. Also, due to lack of tools to measure the user experience from the server-side of VDCs, management functions in VDCs such as configuring thin client protocol parameters are often performed using guesswork, which in turn impacts user QoE. To address the above research challenges in developing scalable VDCs with satisfactory thin-client user QoE, the PIs have developed a ?VMLab? pilot infrastructure that can support desktop virtualization experiments for research and education communities. Over the last two years, this infrastructure has supported: (a) research and development activities relating to VDC resource allocation and thin-client performance benchmarking, (b) desktop virtualization sandboxes for system administrators, (c) virtual desktops for classroom labs involving faculty and students, and (d) evaluation of the feasibility to deploy computationally intensive interactive applications (e.g., remote volume visualization) in virtual desktops. This project will expand the VMLab infrastructure and enable system measurement and user experience evaluation during productive use. Broader ImpactThe broader impacts of this project include serving high-school students and other users from under-represented groups. The proposed research will explore the potential for using the thin-client computing model to provide high-quality user experience to groups otherwise unable to purchase or maintain a full blown PC. The thin-client model may not only be cost effective in terms of purchase costs (assuming end-users do not have to amortize the cost of back-end infrastructure, which is the case in this scenario), but also has a low steady-state cost due to simplicity of device maintenance and troubleshooting. The area of desktop applications hosted in a cloud is an important and rapidly emerging area (both in academia and industry). The broad dissemination of results will have positive impact in an economically valuable technology.