Research >> Projects

Dynamic Schema Evolution in Large Heterogeneous Database Environments

Back to List

T he objective of this work is to create a comprehensive methodology for managing and incorporating changes in database schemas in a heterogeneous distributed environment. A database represents a snapshot of the real world at some point in time. As the real world changes over time so does the database schema. Schema evolution is the ability to incorporate changes in the database schema while preserving existing information in the database. This is an important issue in a heterogeneous database environment. In such an environment, a global or a set of federated schemas may be used to conceptually model the underlying databases. Changes to one or more databases could result in a set of cascading changes to the federated schema. This in turn could trigger changes to other databases. 

It is therefore very important to manage schema evolution which includes identifying and propagating changes to the schema and mapping these back to the set of underlying databases.

We use graph theory to propose a methodology for detecting and incorporating schema changes without affecting existing applications. We hope to automate the process of dynamic schema evolution to the extent possible. A set of software tools that facilitate schema evolution by assisting the database administrator is being prototyped in a UNIX/AIX environment.



For more information, please contact us.