A multiple sparse tables approach for multi-tenant data storage in SaaS

The sparse table approach is an effective storage solution for multi-tenant data in SaaS. However in the sparse table, because of the shared data storage of different tenants that have different schemas, there are so many nulls which result in a waste of storage space and query efficiency problems. We analyze the difference between the multi-tenant sparse data and the traditional sparse one such as e-commerce data and propose a multiple sparse tables approach, in which the tables have gradient columns. We carry out experiments and the results figure out that the approach can effectively reduce the space waste and improve the query performance.