Resumen de Inconsistency management in schema evolution and data base migration
Automation of conceptual schema evolution process should allow database management systems to smoothly perform schema transformation and data migration. However, current approaches lack the ability to perform well in all circumstances. They can’t manage, for example, schema restrictions strengthening without lost of information or the aid of an external migration tool, which, besides of being costly, have to assume database correctness responsibility. In this paper we present a novelty approach to schema evolution automation that performs well in all circumstances. It allows the schema transformation and data migration to be performed in the realm of day by day system use without lost of information or the use of an external migration tool. The approach key idea is to make database management system, not only capable of performing schema changes and data migration, but capable of managing schema evolution/data migration inconsistencies. Instead of pretending an abrupt elimination of evolution related inconsistencies, our approach opts for predict, permit, detect and gradually eliminate them. The approach elements are stated in the framework of the formal approach to the object oriented software development, OASIS. OASIS allows us to rigorously specify and to automatically obtain a software ambient prototype oriented to the creation, animation, evolution, versioning and versions interoperation, of OASIS conceptual schemas, that has, in addition, inconsistencies management capabilities.
