关系数据库中实时高效元数据存储算法优化研究

摘要

The current metadata stored in relational database algorithm not only the required storage space is big,and balance and stability,to read and write inefficient.Therefore,a new algorithm of real-time and efficient metadata stored in relational database,the snake that time slot algorithm based on priority for metadata stored in relational database,by dividing the grid region of a certain type in the relational database metadata stored in the corresponding grid,the metadata stored in the highest priority and query node area,near the first been search highpriority metadata.For priority coil slot algorithm without considering the metadata cache factors,lack of aiming at the disadvantages of metadata access locality optimization mechanism,introducing elements stored surial for improvement,in order to improve the metadata stored real-time efficiency of the algorithm.Experimental results show that the distribution degree of equilibrium and stability of the proposed algorithm is very high,and the storage space is small,read and write performance.%当前关系数据库中元数据存储算法不仅所需存储空间大,而且均衡性和稳定性差,读写效率低下.为此,提出一种新的关系数据库中实时高效元数据存储算法.通过基于优先级的蛇形时隙算法,对关系数据库中的元数据进行存储;通过分割网格区域把关系数据库中某一类型的元数据存储于对应的网格中,把优先级高的元数据存储于和查询节点较近的区域,使得高优先级元数据最先被搜索.针对优先级蛇形时隙算法未考虑元数据存储时缓存因素、缺乏针对元数据访问局部性优化机制的弊端,引入元素存储生存期对其进行改进,以提高元数据存储算法的实时高效性.实验结果表明,所提算法的分布均衡度和稳定性很高,而且存储空间小,读写性能优.