Design Essay

1830 WordsMar 19, 20138 Pages
Design The proposed design will follow a top down implementation as advocated by Inmon. This philosophy recommends building a large centralized enterprise-wide data warehouse followed by several satellite databases to serve the analytical needs of the departments which will become data marts. This approach is predicated on using a relational model for the data warehouse which will attain enterprise wide consistency at a granular level which can later be spawned of into dimensional models in the departmental data marts. A star-schema will be used in each data mart which will store subject specific aggregate data. This approach will ensure the separation of detailed-leveled data from aggregated data. Each data mart with a star schema data model is a relational approximation of the OLAP data model and will be an excellent starting point for building OLAP cube definitions. This will help keep the cost of building OLAP models low and reduce development time and will facilitate Rolap implementation of OLAP cubes and interfaces. Global data warehouse data modeling Inmon believes in creating a relational-model (third normal form: 3NF) Inmon argues that once the data is in a relational model, it will attain the enterprise-wide consistency which makes it easier to spawn-off the data-marts in dimensional-models. Inmon believes the data warehouse data model should be built from the existing corporate data model as a starting point and then modified to conform to a new model that includes all relevant data and includes a dimension of time. [ (Inmon, 2002) ] Inmon believes in a phased approach in which main subject areas are identified and implemented sequentially. The main subject areas in this case will follow from the existing departmental divisions present at the university. This approach will allow us to bring on line specific departments in a serial fashion down to

More about Design Essay

Open Document