Skip to Main Content
Data warehouses are used in many applications that depend on distributed systems. A data warehouse supports information processing by providing a single platform of integrated, historical data for doing analysis. Data warehouses provide the facility for integration in a world of unintegrated application systems. The contents of a data warehouse are evolved in an evolutionary, step-at-a-time fashion. A data warehouse organizes and stores the data needed for informational, analytical processing over a long historical time perspective. Data warehouses keep a materialized view (such as historical data), and user queries are processed using this view. The view has to be maintained to reflect the updates done against the base relations stored at the various distributed data sources. Detecting and extracting modifications from information sources is an integral part of a data warehouse. For unsophisticated sources, in practice it is often necessary to infer modifications by periodically comparing snapshots and backup copies of data from the source. This study considers the materialized view and its maintenance. Various implementation and performance evaluation of the differential snapshot algorithms have been compared for evaluation of suitable alternatives.