ITR: Collaborative Research: Digital Resource Profiling for Wide Area Applications

该提案将探讨影响广域应用程序性能和可伸缩性的基础研究问题。这项拟议的研究对通过动态广域网访问的数字资源随时间和跨不同应用程序的访问延迟和行为变化进行了全面研究。这项研究的一个目标是开发适当的延迟配置文件和资源配置文件来表征这种行为。将开发延迟配置文件，以反映一组客户端和资源(服务器)所经历的端到端延迟。资源配置文件反映了适当的指标，如存在检查、随时间推移的可达性测试以及更新的学习模式。这些配置文件将用于定制向客户提供的服务和信息，同时考虑到应用程序要求和嘈杂的广域网环境。相应的算法将探索端到端延迟和数据过时之间的权衡，以满足不同的应用要求。研究结果将旨在建立一个一致的概况分析框架，并回答以下问题：在多大程度上可以利用概况分析来改善客户获得资源的情况。在理论研究中开发的正式结果将在大规模应用中进行评估。与国家研究计划公司(CNRI)合作处理协议开发商将为大规模提供试验台

This proposal will explore fundamental research issues that impact performance and scalability of Wide Area Applications. The proposed research undertakes a comprehensive study of the access latencies and changing behavior of digital resources over time and across different applications, when accessed via the dynamic WAN. One objective of this research is to develop appropriate latency profiles and resource profiles to characterize this behavior. Latency profiles will be developed to reflect the end-to-end delay experienced by a cluster of clients and resources (servers). Resource profiles reflect appropriate metrics, such as existence checking, testing reachability over time, and learning patterns of updates. These profiles will be used to customize service and information delivery to clients, considering both application requirements and the noisy WAN environment. The corresponding algorithms will explore the trade-offs between end-to-end latency and data obsolescence for different application requirements. Research results will aim to establish a consistent framework for profiling and to answer the following question: to what extent can profiling be used in improving clients' availability to resources. Formal results developed in the theoretical study will be evaluated on large scale applications. Collaboration with Corporation for National Research Initiative (CNRI) Handle protocol developers will provide a test-bed for large scale