Cloud computing can be no longer considered just a hype, as nowadays it is part of the agenda of many companies. Numerous application systems are available as software as a service solutions, however, not only are new developments spurring the market. Yet, traditional software vendors are often confronted with the problem of porting their existing solution into the cloud. In order to analyze and evaluate the process of transformation in advance, a structured approach is required. Holistic knowledge management, combining expertise to develop a SaaS solution and enablement of a targeted access to knowledge, is a basic prerequisite for a meaningful analysis.
The focus of this work is to develop a method, which can be used as a knowledge management tool to evaluate the transformability of on-premise application systems. The basic element of the method, which was developed in cooperation with the msg systems AG, is a data- and filtering structure. In this way solution approaches to fulfill the SaaS criteria, such as the elastic provision or IP-based access, can be classified in a matrix. The dimensions of the matrix can be spanned by directed graphs that specialize the essential criteria to specific requirements, for example scalability under a certain load profile. By specifying his requirements, the user filters relevant paths.
Since cloud computing is an interdisciplinary model to provide IT resources, the method allows considerations of the transformation from different perspectives. Moreover, it provides specific recommendations for action, the feasibility of which must be assessed in the study.
By identifying possible solution approaches, fundamental characteristics for the specification of a SaaS solution could be determined as well. Beside traditional attributes such as load profiles or billing metrics, also the maturity level of scalability, extensibility, or the sharing of resources with others, is crucial.
The method was exemplarily introduced at the cooperation partner. There it shall be further developed and validated. In this first implementation a interdisciplinary applicability could be demonstrated already.