Metadata is a description of data. It provides a consistent way to describe data structures. It is used to describe data as it is being transformed and gives a clear explanation of the meaning of different fields, measures, levels and dimensions. All releases of Microsoft development and data management tools support and allow access to the Repository. The storage of data about data in a central location has some advantages.

Reusability: With the storage at a central location the data can be used and reused.

Dependency tracking: the repository stores information about different objects and the relationships between stored objects. Hence tracking of information regarding relationships is possible.

Tool interoperability : The tools can be used in different systems

Data resource management: Global meta data is stored in the repository for the enterprise data warehouse. It contains a resource library of available services and components, which makes the data easier to manage and more efficient.

Team development: The repository stores meta data about a project’s source control and enables the management of concurrent activities on different versions and configurations of project design and development.

There are two type of metadata in the Analysis services system—Technical metadata and Business metadata.

Technical metadata describes data in a clear and unambiguous way. It contains names of fields, tables, databases, levels, hierarchies and so on. In short it is information about the technical aspects of data.

Business metadata describes data to non technical users so that they can understand the information they are viewing. This includes descriptions of fields, tables, databases and so on.

The Repository technology is the core of SQL Server 2000 data management services. The Repository consists of Information models, the Repository engine, XML interchange, and extensibility.

The Microsoft Repository is a storage place for data about data. It is organized around a set of objects called information models. An information model is a template for a particular kind of data. Information models define meta data in terms of object types and their relationships. The information models have a hierarchical relationship with more specific models inheriting characteristics from their parent models. The repository information helps users assess the impact of changes, aids them in tracking down problems and understand the data they use to make mission critical decisions. The repository is therefore an important part of the data warehousing strategy.

The root model from which all other information models are derived is the Unified Modeling Language(UML). Though users can define their own information models for describing data Microsoft created the Open Information Model(OIM) as a standard specification for storing information about systems and warehouses.

The information model for COM is different from information models of other data types. The relationships between models are specified by the generic model. Microsoft uses the Extensible Markup language (XML) to integrate support for exchanging information in the Repository.

Next Page: Tutorial 23: MSAS : Meta data Repository - Page 2