摘要
网格资源发现算法是网格资源发现机制的核心,带回馈机制的网格资源发现算法充分利用响应报文与请求报文可能经不同路径返回的特点,在正向搜索失败时,启动反向搜索机制,在一个往返时间里进行二次搜索.但是算法在讨论时将整个网格的拓扑结构看成是一个全分布的系统,所有节点位置平等,而现实中的网格大部分是由多个不同的自治系统联结而成,具有一定的区域性.本文对带回馈机制的网格资源发现算法在多域环境中进行了改进,将资源发现过程分为域内发现和域间发现两部分,分别采用不同的资源发现机制,性能分析和模拟实验表明该算法具有良好的可扩展性和较高的性能,更适合真实的网格环境.
Resource discovery is a challenging problem in grid computing because computational resources are large-scale geographically distributed. Resource discovery algorithm with feedback mechanism chooses different path to send request messages and corresponding response messages. A feedback mechanism is added in response message to rediscovery requested resource if the resource can＇t be found in forwarding path. It provides more chances to satisfy the request. But the algorithm treats the grid as a full-distributed environment where every node has the same rank. In fact, real grid is constructed by many different AS. In this paper, we improve the resource discovery algorithm in multi-domain grid. The algorithm uses different mechanisms to deal with different environment such as intra-domain and inter-domain. Performance analysis and simulation results show that the algorithm has good scalability and higher performance. It is more compatible with real grid environment.
出处
《小型微型计算机系统》
CSCD
北大核心
2008年第6期1078-1082,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金重大研究计划(90104005)资助
关键词
回馈机制
网格资源发现算法
多域环境
feedback mechanism,grid resource discovery algorithm,multi-domain environment
