An increasing range of industries, users and society at large have a growing dependence upon software-based control systems, communication systems, applications and services. The key characteristics of these systems are their complexity, large-scale distribution, inclusion of legacy and Off The Shelf (OTS) components, mobility and flexibility. The systems cover safety-, mission-, and security- as well as business-critical systems and a broad range of emerging applications areas such as embedded systems in consumer goods, health support, personal communications, electronic commerce and transport. The novelty of the applications, the extensively deployed nature of new technologies and their common characteristics give rise to new types of problems and challenges to dependability.
Up to date, there is no general methodology for justifying the reliance on the services these systems deliver. The aim of DBench is to contribute to the definition of this general methodology.
DBench will provide to system developers and end-users means for i) assessing the dependability of a system, ii) identifying malfunctioning or weak components of a system which require special attention, iii) tuning a particular component to enhance its dependability, and iv) comparing the dependability of alternative or competing solutions. Therefore, it will help to develop highly dependable systems in an efficient way, thus increasing the competitiveness of system providers.
The DBench project addresses two application areas based on two families of OTS operating systems: embedded systems, and transactional systems with database and web servers, all integrated in network-based infrastructures.
DBench will develop concepts, guidelines and prototypes for dependability benchmarking. It is very important that the same concepts and guidelines be generalised and used in all countries of the EC and abroad, in order to make all systems comparable from the dependability point of view.
Going further, as the purpose of the project is the globalisation of dependability benchmarking, the main guidelines for this purpose can contribute to formulate and implement standards, European or global. DBench will provide useful inputs for the standardisation bodies to define guidelines for the certification of critical systems incorporating OTS components. In this case, the challenge is to reach a global consensus for dependability benchmarking standardisation.
Finally, recent developments in computers and communications enable operation of large systems formed by a very large number of components. The provision of effective conceptual and experimental frameworks for benchmarking system and component dependability to select the most suitable subsystems from the dependability point of view is likely to reduce the significant manpower needed to develop OTS-based systems. It is thus a potential to an achievement to contribute to increase the productivity in these booming service sectors where qualified manpower is scarce.
Project: IST-2000-25425, Action Line IST-2000-5.1.4
Partners: LAAS/CNRS, Chalmers University, University of Coimbra, Friedrich Alexander University, Polytechnical University of Valencia, CRITICAL Software S.A., Microsoft (sponsor).
Industrial Advisory Board: Astrium, CMU, INDRA, Oracle, Saab Ericsson Space, Thales.
Coordinator Contact: Karama Kanoun (
This e-mail address is being protected from spambots. You need JavaScript enabled to view it
), LAAS-CNRS, France