云计算的核心是在虚拟化技术的基础上,通过互联网技术为用户提供动态易扩展的计算资源。利用中心服务器的计算模式来管控网络上大量云资源,使得中心服务器成为整个系统的瓶颈,不利于云计算的大规模应用,因此提出使用对等网络技术构建分布式的云资源索引存储和查询系统,但是结构化拓扑系统维护比较复杂,一般不支持复杂搜索条件查询。提出了一种多关键字云资源搜索算法,在基于分层超级节点的云资源搜索算法基础上进行路由算法改进,希望实现多关键字的精确查询。对多关键字的生成、分割及存储作出了详细说明,提出一种有效的基于数据集的索引搜索策略,实现了包含三个或三个以上的关键字高效、准确查询。分析实验结果证明了算法明显提高了资源搜索的命中率,尤其是随着关键字数目的增多,不仅保证了资源搜索的命中率,同时也大大增加了资源的召回率。
Providing the users with dynamic and scalable computing resources through the Internet technology based on the virtualization technique is the core of the cloud computing. People used to manage a large number of cloud resource with computing model of a central server, which enables it become the bottleneck of the whole system and is not conducive to large scaleapplication of cloud computing. So this paper proposed the distributed cloud resource index storage and query system with P2P technology. However, the topological structure of the system maintenanced more complex, generally did not support complexsearch queries. This paper proposed a new algorithm called a cloud resource search algorithm based on multi-keywords.Improving the routing algorithm based on hierarchical routing supernode search algorithm-based cloud resources, to achieve more precise keyword query. Making a detailed description of multi-keyword generation, segmentation and storage, it proposeda data-set-based index searching strategy to achieve more efficient, accurate query over multi-key searching. The analysis ofthe results shows that: the algorithm is proved effective hit rate significantly increases resources to the search, especially withthe increase in the number of keywords, not only to ensure that the resource search hits, while greatly increasing the recall rate of resources.