Research And Implementation Of Multi-tenancy Data Management System For SAAS Application

With the development of cloud computing technology, software as a service(Software as a service, SaaS), have been proposed as a new software delivery model.SaaS model has enriched application software usage mode. People no longer need toset up their own server, purchase software licenses, deployment and maintenance ofsoftware applications to use the software. In SaaS model, customers use softwarethrough the Internet according their own needs, which is leased from softwareservices provider. SaaS software applications serve multiple tenants by using a singleinstance. In SaaS application, various tenants have their own data model; differenttenants have different business requirement; some-times tenants have the samerequirement for the applications, but there exist the different parameters and processesbetween tenant’s applications.In this paper, shared storage for multi-tenant SaaS application data and businessprocess customization for different tenants are researched; multi-tenancy datamanagement system for SaaS application are designed and developed. We develop theelectronic health records management system for intracerebral hemorrhage patient byusing the multi-tenancy data manager system, and verify the query efficiency throughexperiments. Specific work includes the following sections:Analysis of the data management model in SaaS application developmentprocess, a hierarchical model of data management for multi-tenant SaaS applicationsis proposed. The model shields the concrete realization of the multi-tenant datamanagement by tenant independent data access method. For the concrete data storageway of the multi-tenancy data in the hierarchical model of data management, weanalyze data storage mechanism and the extended model, a data storage method basedon multi-universal data table and tenant data model customized approach based onkey-value pair data model is proposed.In order to satisfy the tenant’s individual requirements, process customizationmethod based on the variability descriptor. We use variability descriptor to describethe configuration section in the application template, and it can be configured byselceting alternative values. Based on the variability descriptor, a generation methodof customization processes model which can guide business customization isproposed.At last, the paper designs and implements the multi-tenant data managementsystem for SaaS application, and we use the system used to develop electronic healthrecords management system intracerebral hemorrhage patient. Experimental resultsshow that the data storage model has a high storage density, can effectively support the customization of the data model of the tenants. The data querying and writing hasa better performance.