Data virtualization is both a data management approach and an enterprise pattern, allowing client applications to access enterprise data without requiring its technical details: format, physical storage, or geographical locations. The main objective of data virtualization is to provide a real-time single view of enterprise data. It differs from various data management paradigms, such as the following ones:
- ETL (Extract Transform Load): With data virtualization, original data source content is not extracted to feed the target client application repository. On the contrary, data sources are kept in place, and only the required data is accessed on demand and in real time. Caching can be used here to improve performances but it is not mandatory requirement.
- Data Federation: Data federation is a type of data virtualization; however, it tends to...