摘要
针对网格领域中资源花费的问题,在分层结构的基础上提出了一种基于综合指数法的网格资源发现及准入策略。该策略描述了一种分层的网格系统结构模型,提高了网格环境下资源发现的效率,并能够综合考虑网格任务执行费用和执行时间两种因素,利用综合指数法对主动声明加入网格的资源状态进行多属性综合评价。在顺利进行资源发现的前提下,针对需要控制花费的网格任务,建立相适应的子网格,使任务计算花费和性能得到最优化。最后,搭建模拟环境表明了该算法的可行性和有效性。
To solve the problem of the enormous expenses of this resource discovery in the field of grid, on the basis of hierarchical structure, a strategy for grid resources discovery and admittance based on comprehensive index method is put forward. The basis of this strategy is a kind of grid model based on hierarchical structure. The strategy synthetically considers two factors: Execution cost and execution time of grid jobs, and conducts multiattribute comprehensive assessment on state of the resources that declare actively to join the gird. On the premise of conducting resources discovery smoothly, this strategy can establish corresponding sub-grid aimed at reducing the expenses of the grid jobs. The results show that the proposed strategy can optimize the computing expenses and capability. Eventually simulation environment is established so that the feasiblity and validity can be validated.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第8期1656-1659,共4页
Computer Engineering and Design
关键词
网格计算
层次
综合指数法
资源发现
集群
grid computing
hierarchy
comprehensive index method
resource discovery
cluster