Documentation and evaluation of data environments for data warehouse

Abstract

In order to support Integrated Operations ConocoPhillips Norge (COPNO) has in cooperation with the “Good to Great” and “Data to Decisions” projects undertaken the task of creating a data warehouse system they have called “Knowledge Infrastructure”. In conjunction with this project COPNO wishes to examine the system and give an overall description of the Knowledge Infrastructure. COPNO also wants to investigate into the possibilities of separating the development- and test environments as well as the chance of creating other environments. In this thesis the system at COPNO is described and evaluated against current problems, especially the issue of separating user data marts from the batch environment and the limitation of only having two environments, namely one “DevTest” environment and a production environment. Further, possible solutions to the environment issue are discussed with focus on best practice from SAP. A more complete solution implements a total separation of the production environment from any other environments. The production environment would then run on hardware separate from the rest of the system. Also, the thesis has arrived at the conclusion that there should exist separate development and test environments. Although bringing more complexity to the overall system, it is also the opinion of this thesis that creation of a sandbox environment could involve advantages and be favorable for the system. Based on the needs and use of the system, several possible solutions to the addressed problem could be presented. An evaluation of costs versus funds and gains to the system needs to be performed in order to make a final decision as to which solution is favorable. Further work consists in implementation of any adopted solutions based on a managerial decision as to whether they have any real value as against the current configuration of the system.