论文部分内容阅读
网格计算(grid computing)是近几年发展起来的一个崭新研究领域,引起国内外学术界及工业界的广泛关注。其目的是研究如何安全有效地将现有的各种计算资源(尤其是那些分布在Internet的异构网络中的计算资源)组织起来协同解决复杂的科学及工程计算问题。在化学信息学和生物信息学中最典型的应用是虚拟高通量筛选侯选药物分子。本文以两个Linux机群为基础,用开放源码的网格支持软件包Globus Toolkit 3.2及Sun~(TM) ONE Grid Engine 5.3成功构建了计算网格;并通过设计测试程序实现一次性提交多个作业(300个)以及分析作业在计算网格中各个节点的分配及运行情况,从而测试了计算网格的效率。结果表明,所构建的计算网格在保持原机群运行稳定、可靠的前提下,改进了系统资源的分配管理方式以及用户提交作业的方法,从整体上提高了网络计算资源的利用率,也同时方便了系统的管理。
Grid computing is a brand new research field that has developed in recent years. It attracts a great deal of attention from academia and industry at home and abroad. Its purpose is to study how to safely and effectively organize various existing computing resources, especially those distributed in the heterogeneous network of the Internet, to jointly solve complex scientific and engineering computational problems. The most typical application in chemical informatics and bioinformatics is virtual high-throughput screening of candidate drug molecules. Based on two Linux clusters, this paper constructs a computing grid with Globus Toolkit 3.2 and Sun ONE Grid Engine 5.3, which are open source grid support packages. By designing the test program to submit multiple jobs at one time (300) as well as analyzing the distribution and operation of each node in the computing grid in order to test the efficiency of the computing grid. The results show that the proposed computational grid improves the allocation and management of system resources and the method of user submitting jobs while maintaining the stable and reliable operation of the original fleet, and improves the utilization rate of network computing resources as a whole Facilitate the management of the system.