An article summarizing results from the EU-Project Q-ImPrESS was accepted for publication in Springer’s Journal on Empirical Software Engineering. It describes how we modelled an industrial control system for performance and reliability predictions. The paper discusses the prediction accuracy as well as the effort for creating the respective models. The paper is based is an extended version of a paper from ISSRE 2010 and another one from ICSE 2011.
This paper shows how our software tools Palladio and PerOpteryx were used in capacity planning for a web-based remote diagnostics system at ABB. It will be presented at the International Conference on Performance Engineering (ICPE 2012) in April in Boston, US.
PerOpteryx results: CPU Utilization vs. costs; each dot represents a generated design alternative for the system. The diagram shows the trade-off between performance and costs.
This is my talk at the ICSE 2011 Software Engineering In Practice Track. It’s about a large-scale case study on performance and reliability prediction for service-oriented software systems. Also see the Q-ImPrESS website for more details.