Wits Technikon, Medunsa and Technikon Pretoria have joined forces on a data warehousing project using SAS Institute`s decision management architecture to supply management information systems to administrative and teaching staff.
The project is being managed from Wits Technikon where the entire data warehouse will be housed on an Orca 3000 attached to an HP 9000 D350 dual processor, all supplied by Computer Configurations.
Data marts are being created in four areas: students, finances, human resources and space management. The student data mart enables staff to access information such as which courses a student is registered for and marks obtained in exams. Administration staff members can see how many students are registered for a particular course and pick up trends over periods. This in turn enables the management of human resources and accommodation. The student data mart has been completed and work recently began on the financials mart.
"Our main objective is to empower our end-users," says Stephan Botha, head of MIS. "They`ll be able to access the data marts via Web browsers. In fact, we chose SAS because of its ability to integrate into the Web."
At Wits Technikon all staff members who are connected to the intranet can access the data warehouse. SAS/MDDB Server software is being used to create multidimensional databases for online analytical processing (OLAP).
The Orca 3000 has been bought to manage the mass storage required by the new data warehouse.
"Because data warehousing is resource-hungry," says Botha, "we decided on the Orca because of its scalability. Its 40 Gigabytes of RAID can easily be trebled. The Orca`s configuration of disks to controllers also increases throughput speeds and response times."
"Wits Technikon`s purchase of the Orca fits in with their practical growth path, enabling them to keep abreast with technology, yet maximise the efficiency of their systems in a cost-effective way," says Jim Jeffrey, sales director for Computer Configurations. "The Orca`s scalability enables them to grow according to their needs as the data warehouse project moves through its development stages."
Share
Editorial contacts