Abstract
The data integration systems under large-scale, distributed environment require more flexible and scalable. The existing approaches based on multi-database cannot meet this requirement. This paper presents a CORBA/IIOP-based plug and play service model underlying data source systems, and the implementation issues in the CORBA-based information integration system SCOPE are also discussed, including the wrapper technique and the reference definition for distributed objects. According to the requirements of applications, two kinds of objects including volatile objects and persistent objects are proposed, and the associated join-in and withdraw techniques are discussed in detail. The plug and play mechanism for data integration avoids coding for each registering data source. A join-in or withdraw processing will not disturb other distributed objects. Moreover, the schema extracting service is provided to support information inter-operation and integration. The application shows that the mechanism has advantages of flexibility, scalability, and feasibility for distributed information sharing systems.
| Original language | English |
|---|---|
| Pages (from-to) | 341-347 |
| Number of pages | 7 |
| Journal | Ruan Jian Xue Bao/Journal of Software |
| Volume | 12 |
| Issue number | SUPPL. |
| Publication status | Published - Jun 2001 |
| Externally published | Yes |
Keywords
- CORBA
- Data integration
- Database
- Multidatabase system
- Software bus