Keywords object-oriented databases, relational databases, SQL
Start Date: 01-NOV-93 / Duration: 36 months
[ contact / participants ]
The IRO-DB project intends to develop a set of tools to achieve interoperability of pre-existing databases and new object-oriented databases. It will concentrate on providing a C++ library to access heterogeneous databases supported by communication protocols to exchange object-oriented SQL commands and objects.
This library will be an extension of the object-oriented world of evolving (de facto) standards, ie, the Call Level Interface of the SQL Access Group. Furthermore, tools to design and dynamically maintain integrated applications on large federations of heterogeneous databases will also be developed. Together, the developed tools will constitute a Federated Object-Oriented Database Systems mediating between heterogeneous databases and integrated applications. They will provide uniform C++ interfaces on top of dynamically adaptable integrated schemas for data exchange and sharing, global transaction management and concurrency control, without violating the autonomy of the participating databases.
The first phase of the project will be devoted to developing a simplified version of the system for querying distributed relations and classes with mono-valued attributes and references on the basis of export schemas. The object interchange protocol will be designed as an object-oriented specialisation of the generic remote database access (RDA) standard, considering the format and protocol (FAP) proposed specialisation of the SQL Access Group. It will support, in addition to the main features of the already available specialisation of RDA for relational databases (SQL RDQ), classes of complex objects possibly interrelated using object identifiers, and generalisation hierarchies. To support this protocol (called OORDA) operationally, communication software and procedures for interfacing local databases will be implemented on top of INGRES and MATISSE. On top of the OORDA protocol, a C++ library, and a first version of a schema integrator with a query translator will be provided as attractive tools to develop object-oriented applications using existing DBMSs.
The second phase will mainly extend the tools developed in the first phase and interconnect other database servers (O2, ONTOS). The OORDA protocol will be extended with a rich set of type constructors, behaviour encapsulation, update and facilities to maintain atomic and concurrent transactions. The schema integrator will be extended to incrementally and co-operatively design and maintain integrated schemas for multiple groups of users. More powerful mapping facilities between local database schemas and export schemas will also be integrated in the system. Integrated schemas will be operationally supported by a more sophisticated interoperable object manager, providing for processing optimised queries over multiple sites and for global transaction management and concurrency controls.
One of the main results of the project will be the provision of an easy path for integrating the relational database technology to the new object-oriented one. Thus, most existing applications willing to interoperate with existing databases and object-oriented approaches should benefit from the project. Within this project, a demonstrator will be built to evaluate the system using an engineering application. The application will be abstracted from an existing one at INTRACOM, the major telecommunication company in Greece. It will integrate an existing CIMr database running on INGRES with an SQL production database and an abstraction of a CAD database developed on top of an object-oriented system (in principle, MATISSE). Further investigations of applications, notably in the banking domain, will be done by project partners (IBERMATICA in particular).
Due to its layered design and its strong orientation towards evolving standards, the project will also result in two software products. These are a remote database access link for programming object-oriented applications in C++ on top of various database systems using a call level interface extending the CLI interface promoted by the SQL Access Group, and a schema integration and query translation tool for developing and maintaining interoperable applications on pre-existing heterogeneous databases. The participants in the consortium, which are mainly DBMS providers and software developers, will all make special efforts to exploit the resulting products, both as independent tools and as components integrated within applications.
Mrs Maria Dolores Garcia Plaza
Avenida De Burgos 9
E - 28036 MADRID
tel: + 34/ 1-3849100
fax: + 34/ 1-7671565
IBERMATICA - E - M
GMD - D - P
GOPAS - D - P
INFOS - F - P
IN&LITIC - F - P
INTRASOFT - GR - P
O2 - F - P
IRO-DB - 8629, December 1993
please address enquiries to the ESPRIT Information Desk
html version of synopsis by Nick Cook