Model-driven architectural monitoring and adaptation for autonomic systems
ICAC '09 Proceedings of the 6th international conference on Autonomic computing
Adaptation and abstract runtime models
Proceedings of the 2010 ICSE Workshop on Software Engineering for Adaptive and Self-Managing Systems
Incremental model synchronization for efficient run-time monitoring
MODELS'09 Proceedings of the 2009 international conference on Models in Software Engineering
Achieving dynamic adaptation via management and interpretation of runtime models
Journal of Systems and Software
| Hi-index | 0.00 |
During the last decades, performance of available hardware resources constantly increased [15], which enabled the assignment of more and more complex tasks to software systems. As one consequence, the inherent complexity of these software systems also increases, influencing all phases of their lifecycle. The concept of Component Orientation (CO) [17] allows the development of software systems in a modular way through functional decomposition. Administration and maintenance of software systems are addressed by the vision of Autonomic Computing (AC) [11], based on the idea to assign low level administrative tasks to the system itself. With mKernel an AC-infrastructure for component oriented enterprise applications is provided, based on the Enterprise Java Beans (EJB) standard, version 3.0 [5]. In contrast to existing approaches, the main advantage of mKernel lies within its standard compliance, not prescribing any additional guidelines for the development of applications to enable their autonomous management. It is realized as plugin for an existing container, not requiring any adjustment of the underlying implementation. Moreover, it provides a very fine grained interface for inspection and manipulation of the managed system, taking the specifics of the supported standard into account. Within this paper we present the opportunities provided by mKernel to control a managed system.