An overview of WIND (Wide Interoperable Networked Databases)

This paper gives an overview of the ongoing project WIND (Wide Interoperable Networked Databases), a multidatabase management system integrating legacy databases distributed on a network in a transparent fashion. WIND provides users with a relational view over a number of local databases and SQL as the data sublanguage facility. WIND supports multiple federated schemas in the sense that users with similar information requirements are clustered around the same relational view. Operationally, there are three ways of maintaining the multidatabase: transactions (preserving ACID properties), procedures (weak ACID database applications) and event-trigger rules. At the data layer level, WIND supports a wide class of interdependent data comprising base relations, materialised views (snapshots) and replicated data.<>