Next Generation Software: Logistical QoS through Application-Driven Scheduling of Remote Storage

EIA-9975015 Plank，James S.University of Tennessee-Knoxville下一代软件：通过应用驱动的远程调度实现Logistics QoS（Logistics Computing and Internetworking - LoCI）旨在开发能够提供可靠、灵活可扩展且具有成本效益的数据交付服务质量。该提案旨在开发新的方法来实现这些功能，其中包括拟议的后勤QoS，以提供对消息缓冲的灵活得多的控制，而不需要困难的端到端保证要求。 这可以通过开发数据分级等技术来实现，通过提前移动数据并将其存储在接收器附近，拟议的项目将研究物流网络计算的这些概念。该提案将开发这些方法和技术，并使用现有的工具和协议来实现和测试这些想法，例如Internet Backplane Protocol，Network Weather Service，NetSolve，以及PI和他的合作PI多年来开发的应用级调度系统。这一努力将使中间件支持更容易被应用程序利用。

EIA-9975015 Plank, James S.University of Tennessee-KnoxvilleNext Generation Software: Logistical QoS through Application-driven Scheduling of Remote StorageThe research in the proposed project (Logistical Computing and Internetworking - LoCI) is aimed at developing middleware that can provide reliable, flexible, scalable and cost-effective delivery of data with quality of service.The proposal aims to develop new approaches to enable these capabilities, among which is the proposed logistical QoS, to provide much more flexible control of message buffering without requiring difficult end-to-end guarantee requirements. This can be accomplished by developing techniques like data staging, by moving data in advance and storing them close to the receiver, and the proposed project will investigate such notions of the logistical network computing.The proposal will develop such approaches and techniques and use existing tools and protocols to implement and test these ideas, such as the Internet Backplane Protocol, the Network Weather Service, NetSolve, and the Application Level Scheduling systems that the PI and his co-PI's have developed over the years. The effort will result in middleware support more readily exploited by the applications.