Collaborative Research: Resource and Data Management for Virtualized End-resources in Computational Grids

这是西北大学和佛罗里达大学之间的合作建议，它解决了与动态虚拟机（VM）的管理和现有的网格中间件技术与这些机制接口的关键研究挑战。 该项目将开发新的解决方案，解决-分布式虚拟化终端资源的资源管理，可以动态创建的资源，-映像管理，用于按需传输代表整个状态的数据，以创建动态VM实例，和数据管理，用户和应用程序数据在网格上解耦的计算服务器和数据服务器之间的需求传输。这些解决方案将构成网格计算的信息处理基础。 在这个项目中产生的软件将被起诉，以实现下一代网络计算中心，目前正在使用，以支持在纳米技术，电子CAD，计算机体系结构和并行编程。该项目解决了关键的研究挑战，允许管理虚拟机中间件的动态实例，并将导致解决方案，将形成网格计算的信息处理基础，虚拟化终端资源。 特别是，在这个项目中产生的软件将用于实现下一代网络计算中心，目前用于支持纳米技术，电子CAD，计算机体系结构和并行编程的模拟需求。

This is a Collaborative Proposal between Northwestern University and University of Florida which addresses the key research challenges associated with management of dynamic Virtual Machines (VM) and interfacing these mechanisms with existing grid middleware techniques. The project will develop novel solutions that will address- resource management for distributed virtualized end-resources that can be created dynamically,- image management for the on-demand transfer of data representing the entire state to create a dynamic VM instance, and- data management for the on-demand transfer of user and application data between decoupled compute and data servers on the grid.The proposed approach will lead to middleware solutions that will form an information processing foundation for grid computing. The software generated in this project will be sued to implement the next generation of network computing hubs currently being used to support simulation needs in nanotechnology, electronics CAD, computer architecture and parallel programming.The project addresses key research challenges to allow the management of dynamic instances of virtual machine middleware and will lead to solutions that will form an information processing foundation for grid computing on virtualized end-resources. In particular, the software generated in this project will be used to implement the next generation of network computing hubs currently being used to support simulation needs in nanotechnology, electronics CAD, computer architecture and parallel programming.