作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2006, Vol. 32 ›› Issue (15): 55-57. doi: 10.3969/j.issn.1000-3428.2006.15.020

• 网络与通信 • 上一篇    下一篇

基于双数据目录存储的网格元数据的资源发现

孟宪虎1, 2   

  1. 1. 浙江财经学院信息管理系,杭州 310012;2. 山西运城学院计算机系,运城 044000
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2006-08-05 发布日期:2006-08-05

Looking for Resource Based on Grid Metadata Stored in Double Data Directory

MENG Xianhu1, 2   

  1. 1. College of Information Management, Zhejiang University of Finance and Economics, Hongzhou 310012; 2. Department of Computer, Yuncheng University, Yuncheng 044000
  • Received:1900-01-01 Revised:1900-01-01 Online:2006-08-05 Published:2006-08-05

摘要:

提出了基于全局命名的数据网格环境下元数据目录和资源信息目录双目录存储和访问的概念,研究了全局命名的规则,给出了在该命名情况下元数据的注册和存储以及利用这种注册存储如何快速发现远程数据资源。说明了在多网格结点下寻找“对照表”采取时间戳抑制多次重发和导致多次访问数据资源的方法,并指出了“对照表”应采取的热点数据优先、冷僻数据清除的策略。

关键词: 数据网格, 元数据, 双目录存储管理, 全局命名, 数据发现

Abstract: This paper brings forward the concept of storing and accessing data both in metadata directory and resource information directory based on the condition of overall situation naming, then studies the rule of overall situation naming and gives the registering and storing of the metadata. This paper also studies the problem of how to use the register and storage to find the long-distance data resource. Then, it illuminates the method of searching for the “comparison table” using the time marker in the multiple-grid-node. This method can restrain the instance of sending the same message time after time and the instance of accessing the data resource time after time. At last, the paper gives the strategy of giving priority to the common data and cleaning out the rare data.

Key words: Data grid, Metadata, Double directory lists storage & management, Global naming, Data discovery

中图分类号: