Self-awareness: The Foundation for Resilient Computing
The context of changes affecting a system is fundamental for achieving resilience. The contextual awareness should be dynamic in terms of what, how and when to observe and analyse. For example, which contextual attributes to observe may depend on a wide range of historical and current run-time factors. System self-awareness stems from its ability of dynamically building models of the context of the change affecting the system. These models provide the basis for the system to react to change, for instance, by adapting its structure and/or the services it provides. For handling uncertainty, it is not enough to react to the perceived change, the context in which change has happened should also be considered. The provision of this dynamism depends on the system agility when dealing with constraints and opportunities. A promising way for achieving this is through specialised micro-services that can be deployed and reconfigured depending on the perceived system needs. Self-awareness is not restricted to services provided by the system. The provision of quality of services, like security, should also rely on the context of changes affecting security. For instances, in authorisation systems, the operational profile of subjects and resources can influence how privileges are managed.
Rogério de Lemos is a Senior Lecturer in the School of Computing at the University of Kent since 1999. In the past, he was a Senior Research Associate at the Centre for Software Reliability (CSR) at the University of Newcastle upon Tyne, and invited assistant professor at the University of Coimbra in Portugal. His research interests are on self-adaptive software systems, architecting dependable and secure systems, and resilient AI.