Research And Design Of Data Storage Model And Schema Extension Techniques In SaaS

SaaS (Software as a Service) is a new software model which arose at the beginning of this century. It is different from the traditional software in service mode, which developed only for the specific customer (user). SaaS service provider have to provide all network infrastructure and hardware and software operating platform which needed in the users’ information construction to the user, and is responsible for all the pre-implementation and the latter part of the maintenance work of SaaS Services, this will not only significantly reduce the overall investment cost of the construction of enterprise information, but also enable enterprises to more focus on the business objectives of high-value actives.SaaS has so many outstanding features, such as services online, flexibility, functionality can be configured and attributes can be extended and so on, and it can also be self-configured, business process reengineered and service integrated as the software complex and changing demand. SaaS can make the software easy to use, but at the same time, itself also has many of the key technical issues which needed to be resolved. In this paper, we focus our attention on the data storage model and schema extension techniques.The main work of this paper is as follows:Firstly, after in-depth study of the three traditional data storage model, in order to overcome the disadvantages in the models of isolated database and shared database isolated schema, such as too much redundancy field in the definition of product information, the level of sharing resource is too low, etc, and the shortcomings of shared database shared schema, such as the isolation level is too low between tenants and data is not very safe, this paper introduced a new data storage model:shared database semi-shared semi-isolated schema data storage model. It is better than shared database shared schema in the level of data isolation between tenants, and it is also better than shared database isolated schema in the level of share hardware resource among tenants, and we also gave details of its implementation process and related concepts. And this paper compared the four data storage model in the end of chapter three.Meanwhile, this paper introduced six main model extension techniques in the development of SaaS services, and gave details of their implementation process, and also discussed their advantages and disadvantages. And then propose an improved version schema extension technique based on Sparse Columns technique and Universal Table technique. The improved version schema extension overcome the disadvantages of Sparse Columns technique and Universal Table technique, such as the data table is too wide, NULL value is too much and waste of storage space, etc. And at last this paper compared the improved version schema extension technique with the six schema extension techniques mentioned in the beginning of chapter four, so that all of us can clearly understand the advantages and disadvantages of the new schema extension technique in detail.Finally, this paper compared the new data storage model with shared database shared schema and shared database isolated schema based the improved version schema extension technique in service performance by experiments, not only proved the availability of the new model but also showed the effect of its service, and provide a reference for future related techniques progress, and also give the SaaS provider more choice when they develop SaaS service.