Research And Implementation Of P2P-Based SaaS Data Integration

Along with the raising level of enterprise information and growing business collaboration, the demand for information sharing and data interoperability between enterprise applications is increasing. And the applications may be between different enterprises or within a single company. Data integration is an effective way to share information and it is the basis of achieving the interoperability of business data. Therefore, how to efficiently achieve data integration between different applications becomes an important issue in information-sharing field.Based on the research of enterprise information and P2P technology, the paper presents a ECGP-based data integration model referencing the ECGP (An Efficient Clustered Global-Peer P2P Architecture) structure. Compared with traditional P2P-based data integration models, the model is scalable and load balancing. And the model overcomes the low efficiency and managing difficulties caused by excessive enterprise data nodes.Secondly, based on the data integration model, SOA (Service-Oriented Architecture) and SaaS (Software as a Service) model, the paper presents a SaaS-service-based data integration framework and it has the following characteristics:1) The framework is service-oriented:Data sources distribute in different application systems within a single company or different companies, so data integration has a distributed, heterogeneous and autonomous characteristic. Service-oriented architecture, with loosely coupled and platform independent features, could maintain the autonomy of data sources and easily mask the underlying heterogeneous data sources.2) The SaaS model is applied:SaaS model, along with challenges, brings new solutions for enterprise. SaaS-based framework could make data integration between system and SaaS applications or system and non-SaaS applications more efficient and flexible. Besides, the SaaS model could prevent company from deploying, upgrading and maintaining the system, which could save cost for enterprises.3) The combination of data warehouse and middleware is applied:Data warehouse is used to store, manage, and analyze historical data and middleware is used to integrate data inserted, modified or deleted in data sources, which could ensure accurate and complete integration and make a real-time and efficient data integration.Firstly, the paper introduces SBM (Supply Business Management) and ERP as an example, analyzing the business scenario and demands for enterprise data integration, and presents the ECGP-based data integration model. Secondly, based on the above model, the paper introduces SaaS-service-based framework of data integration. And, the frame architecture, integration mechanism, process and modes are described in detail, and then the research on heterogeneity and consistency in data integration is analyzed. Next, the implementation process of system is demonstrated by describing the realization of key modules and the practical application of system is also described. Finally, we summarize the paper and prospect the future...