Guest Editor's Introduction: Creating Robust Software through Self-Adaptation
IEEE Intelligent Systems
The Vision of Autonomic Computing
Computer
A Fault Detection Service for Wide Area Distributed Computations
HPDC '98 Proceedings of the 7th IEEE International Symposium on High Performance Distributed Computing
An Externalized Infrastructure for Self-Healing Systems
WICSA '04 Proceedings of the Fourth Working IEEE/IFIP Conference on Software Architecture
Towards a rule model for self-adaptive software
ACM SIGSOFT Software Engineering Notes
Software Dependability Analysis Methodology
ICCSA '09 Proceedings of the International Conference on Computational Science and Its Applications: Part II
| Hi-index | 0.00 |
Much effort is required to manage complex computing environments. Self-healing systems, adapted to their computing environments, offer a solution to this problem. Additional resources are required to use these systems. In our previous study, we proposed an approach that controls the behaviors of the self-adaptive components. This approach used switches to reduce the amount of resources required and enhance system performance. However, it did not include an abstraction approach to control the behavior of the self-adaptive modules. Specific code needs to be included when designing and implementing self-adaptive modules to utilize abstraction. In this study, we propose a method of 1) analyzing the behavioral levels of the self-adaptive modules through a goal graph 2) generating a behavior level activation switch using the specified levels. Through this approach, we can reduce additional resources required by the self-adaptive modules, and the effort involved in creating the code required for improving their performance. In order to evaluate the proposed approach, we draw a goal graph of a file transfer module used in a video conference system. This approach generates the template code of the behavior switch based on the goal graph that can improve the performance of self-adaptive modules. Through these processes, the generated code is applied to the modules, while verifying that the switches function properly. We compare the number of concurrent active components before and after applying this approach. In this way, we make certain that the code generated from the goal graph improves the performance of self-adaptive modules.