面向SaaS应用基于多宽表模式的多租户索引研究  

RESEARCH ON MULTI-TENANT INDEX BASED ON MULTI-WIDE TABLE MODE FOR SAAS APPLICATIONS

作　　者：张雅文 刘春霞[1] 党伟超[1] 白尚旺[1] Zhang Yawen,Liu Chunxia,Dang Weichao,Bai Shangwang(Department of Computer Science and Technology, Taiyuan University of Science and Technology, Taiyuan 030024, Shanxi,China)

机构地区：[1]太原科技大学计算机科学与技术学院,山西太原030024

出　　处：《计算机应用与软件》2018年第7期63-68,共6页Computer Applications and Software

基　　金：山西省重点研发计划(一般)工业项目(201703D121042-1)

摘　　要：租户按需定制和高效存储是SaaS应用多租户数据存储的两个重要方面,因此需要提供数据存储机制及索引机制。多宽表的数据存储模式中,同一数据列有多种数据类型,基于列数据库索引技术无法应用于多宽表数据存储模式。它给出改进的Indexes Pivot Table索引模型,该模型根据租户定制需求,为各租户业务数据形成索引元数据,利用元数据驱动实现不同租户索引数据的定制和隔离。给出索引维护策略,保证租户索引数据随业务数据同步更新,以及索引机制下数据查询算法。实验结果表明,在两级索引即租户逻辑索引和关系数据库物理索引的基础上,该索引模型对快速访问数据具有较好的性能。Tenant on-demand customization and efficient storage are two important aspects of Saa S application multitenant data storage,so it is necessary to provide data storage mechanisms and indexing mechanisms. In the data storage mode of the multi-wide tables,there are multiple types of data in the same data column. The data storage mode of multiple universal tables is not applicable based on column database indexing technology. This paper presented an improved Indexes Pivot Table indexing model. According to the tenant customization requirements,this model formed index metadata for each tenant business data,and used metadata to drive customization and isolation of different tenant index data. The index maintenance strategy was given to ensure that the tenant index data was updated synchronously with the business data and the data query algorithm under the index mechanism. The experimental results show that,based on the two-level index,that is,the logical index of the tenant and the physical index of the relational database,the index model has good performance for fast access to data.

关 键 词：SAAS应用 多租户 多宽表 存储模式 索引 维护策略 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]