Reliability assessment and prediction of open source software systems

Affordability, redistributable, modifiability, availability of source code, zero price and freedom of choice are the characteristics which have made Open source (OS) a preferred platform for many software industries and individuals, who believe to use the power of high quality software. The technological motivation for OSS development directly relates to the software crisis, which clearly specifies that traditional development do not perform very effectively, specifically in the areas of speed, quality and cost of development. In this paper fact regarding open source software as well as different reliability concepts are elaborated. Practical implementation of OSS reliability modeling and decision making about product release has been done. For this purpose software performance measures Failure intensity, Mean Time to Failure (MTTF) and Reliability of three different open source software´s failure data sets are computed and analyzed. Comparative selection of best suited Software Reliability Growth Model (SRGM) among four SRGMs for concerned data sets have been done on the basis of two comparison parameters Bias and SSE.