An Efficient Approach for Supporting Multi-Tenancy Schema Inheritance in RDBMS for SaaS

Abstract

Multi-tenant data management is a major application of Software as a Service (SaaS) , whereby a third party service provider hosts databases as a service and provides its customers with needed services. SaaS applications are deployed on a shared environment that can be accessed by the users from client-end software by using the Internet. Multi-tenancy refers to a principle in software architecture where a single instance of the software runs on a server, serving multiple client organizations (tenants). Multi-tenant applications provide a common user interface (UI) for all the organizations and data of multiple tenants are saved in a single database to reduce total cost of ownership. Common practice is to map multiple single-tenant logical schemas in the application to one Multi-tenant physical schema in the database. Such mappings are challenging to create .This is due to the flexibility of a base scheme to be extended by enterprise application tenants which provides different dynamically modified versions of the application. The fundamental limitation on scalability of this approach is the number of tables of database can handle. Shared Tables Shared Instances (STSI) is a state-of-the-art approach to design the schema. However, they suffer from performance time and high space overhead. In this research, we are going to introduce an efficient approach for supporting Multi-tenancy schema inheritance based on STSI, that allows sharing core application schema between tenants while enabling schema extensions per tenant, Schema inheritance allows deriving a schema from another schema. Thereby, a derived schema inherits the objects that are defined in the parent schema. The idea is based on the changes that occur at runtime in the meta data and the data . Also exploitation some situations of data needs to be shared between tenants. Several experiments were conducted to trade-off STSI and our approach. Different sizes of small, medium, large and very large databases, starting from 10 GB up to 300 GB. Experimental results show that our method achieves good scalability and high performance with low space requirement, and outperforms STSI methods at different rates depending on data manipulation language (DML) operations. It is ranged 50% in selection processes, and have been in the range of 20% and 40% in the update and insert. Also, our approach achieved less storage space compared with STSI about 50% .