The initial objective of the work presented in this thesis was to contribute a viable alternative to existing cost allocation methodologies (see section 1.2). To evaluate the “viability” of the approach presented in the previous chapters we conducted a proof of concept in cooperation with our industrial partner, the BMW Group. The project was organized into three stages. First, we analyzed the cost allocation practices for IT infrastructure services and verified whether the situation is in accord with our general assumptions on IT Cost Accounting and Chargeback (see chapter 2). Second, by example of Java/J2EE-based application systems we examined how the determination of resource profiles could be integrated into the existing IT Service Management processes. Finally, we applied method and software to two different application systems in a data center of the BMW Group. Results from experiments on that infrastructure were already presented in chapters 4 and 5. During the proof of concept, we put a particular analysis focus on the following two requirements (see section 3.4):
  1. 1.

    Consistency This requirement means that the estimation should be applicable to various IT infrastructures, without a need to change the respective systems. We deliberately chose Java/J2EE technologies for the proof of concept, as the respective guidelines at the BMW Group facilitate the design of distributed and heterogeneous software and hardware architectures. So, we combined in the test infrastructure three different operating systems (Linux, Windows, Unix) with server software from Apache, Bea and Oracle. All hardware and software components were typical enterprise products. By means of this rather heterogeneous, but realistic, infrastructure, we tried to gather significant results on the consistency of the estimation process and the determined resource profiles.

  2. 2.

    Operating Efficiency The determination of resource profiles should integrate well with existing IT Service Management processes and cause little extra work. We addressed this requirement with a feasibility study with Java/J2EE-based application systems. We first developed an integration into existing processes and then conducted a number of test series with two example application systems. Thereby, we tried to adhere to all relevant process and architectural guidelines to get sound estimates for the expected aborts and the operating efficiency of the approach for the determination of resource profiles.



Resource Consumption Business Unit Cost Allocation Hardware System Capacity Planning 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Betriebswirtschaftlicher Verlag Dr. Th. Gabler | GWV Fachverlage GmbH, Wiesbaden 2008

Personalised recommendations