Modelling the Dynamic Reconfiguration of Application Topologies, Faults Included
Abstract
Fault-aware management protocols permit modelling the management of application components (including potential faults) and analysing the management behaviour of a multi-component application. The analysis is driven by the application topology, and it assumes many-to-1 dependencies among application components, i.e. each requirement of a component can be satisfied by exactly one other component.In this paper we extend fault-aware management protocols to account for many-to-many dependencies among components, i.e. different application components can be used to satisfy a requirement of another component. The extension also accounts for dynamic changes in the topology, hence enabling the analysis of the management behaviour of dynamically reconfigurable multi-component applications.
Origin | Files produced by the author(s) |
---|
Loading...