Evaluation of Dependable Layered Systems with Fault Management Architecture

The need for a separate fault-management system, that is able to carry out both failure detection and reconfiguration, is becoming imperative due to the increasing complexity of fault-tolerant distributed applications. Such practice would eliminate the intricacies of the failure detection mechanisms from the application and would avoid repeating them in every program. The dependability of such an application depends on the interconnection of components in the faultmanagement system, management subsystem failures, delays incurred due to system reconfiguration and failure information propagation in the management architecture, as well as on the structure of the application itself. This position paper describes avenues for evaluating the dependability of a multi-layered service system that uses a separate fault-management architecture.