论文部分内容阅读
为了更好地管理和应用数据网格中大量分布异构的数据,在对网格技术发展现状进行深入研究基础上,提出基于数据空间概念的网格数据的管理架构.在此基础上,实现了一个网格数据空间管理原型系统,系统中将现有的一些数据空间技术如数据模型iDM、查询语言iTrails等与网格数据访问中间件OGSA-DAI相结合,使用OGSA-DAI工作流来完成数据空间管理系统的数据访问、抽取、数据索引、服务管理和查询回复等一系列工作.实验表明数据空间管理机制在网格环境下是可行的,数据空间管理系统屏蔽了网格数据的分布性和异构性,且能够适应网格数据动态特性,因此满足了对网格数据的管理要求.所提出的网格数据空间架构为网格数据管理提出了新的方法.
In order to better manage and apply a large number of distributed heterogeneous data in data grid, based on the deep research on the current status of grid technology, this paper proposes a grid data management framework based on the concept of data space. On this basis, A grid data space management prototype system, the system will be some of the existing data space technologies such as data model iDM, query language iTrails and grid data access middleware OGSA-DAI combined with OGSA-DAI workflow to complete Data space management system data access, extraction, data indexing, service management and query response and so on.Experiments show that data space management mechanism is feasible in grid environment and data space management system shields the distribution of grid data And heterogeneity, and can adapt to the dynamic characteristics of grid data, thus meeting the management requirements of grid data.The proposed grid data space architecture presents a new method for grid data management.