Datawarehousing/Plato - Page 16

25 Jan 98

Data Warehousing

Goal: "Make it easier to design, build, manage and use cost-effective data warehousing solutions using Microsoft SQL Server."

"Data warehousing is database made ubiquitous"

The data warehousing efforts use the Microsoft Repository. This will support a variety of information models (UML, CDE, etc.) This is an extension of the repository that comes with Visual Studio. It has been extended to support database schema, transformations and OLAP.

The OLAP server is code-named "Plato". [In this writeup I am assuming knowledge of multi-dimensional processing] It supports the different types of storage used for multi-dimensional processing of data: ROLAP (data stays in relational data store), MOLAP (data is converted to multi-dimensional data store) and HOLAP (combination of ROLAP/MOLAP).

It has lots of wizards. One of these determines when pre-aggregation of data is going to have a payoff.

Plato security will be integrated with SQL Server in Beta 3. Plato is integrated with office and back office.

There are some third party clients already for Plato: Arbor, Knosys and Panorama.

OLE DB For OLAP extensions support this kind of processing. There is a new object called dataset which is the result of querying a cube or set of cubes.

There is a subset of Plato called Data Cube Server which is an in process desktop multi-dimensional component.