Multiversion data warehouses: challenges and solutions

A multiversion data warehouse (MVDW) is a promising approach to handling structural changes in external data sources as well as for simulating alternative business solutions. A MVDW is composed of the set of its versions, each of which is in turn composed of a schema version and an instance version. A single version stores a data warehouse state valid in a certain time period or a simulation scenario. In order to manage and analyze multiversion data, new challenging issues are to be solved, that include among others metadata management, schema and data versions management, data sharing, multiversion query languages, and a transactional DW management. These issues are briefly discussed in the paper based on our experience in designing and implementing a multiversion data warehouse management system.