Towards Systematic Design of Adaptive Fault Tolerant Systems
Elena Troubitsynå, Kashif Javed · 2014
Abstract—The development of modern distributed software systems poses a significant engineering challenge. The system architecture should exhibit plasticity and high degree of reconfigurability to enable an automated adaptation to continuously changing operating conditions and component failures. Traditional engineering approaches are inefficient to cope with complexity of such systems to ensure their robustness and fault tolerance. Therefore, there is a clear need for the approaches explicitly addressing the problem of designing adaptive fault tolerance mechanisms. In this paper, we propose a systematic approach to the development of adaptive fault tolerant systems. We discuss the main principles of architecting such systems to enable plasticity and reconfigurability. We demonstrate how deployment of the predictive adaptation allows us to ensure that the system would be able to continuously deliver its services with the acceptable quality despite occurrence of component failures. Keywords-adaptable systems; fault tolerance, predictive adaptation; reconfiguration. I.