Self-aware QoS management in virtualized infrastructures
Proceedings of the 8th ACM international conference on Autonomic computing
Automated extraction of architecture-level performance models of distributed component-based systems
ASE '11 Proceedings of the 2011 26th IEEE/ACM International Conference on Automated Software Engineering
Fmeter: extracting indexable low-level system signatures by counting kernel function calls
Proceedings of the 13th International Middleware Conference
Communications of the ACM
Queue - Distributed Computing
Hi-index | 0.02 |
The wide-spread interest in self-management reflects the disturbing fact that scaling information systems is often impaired by the burdens of management and operations. These concerns have motivated the development of technologies for self-healing, self-configuration, self-optimization, and self-protection. Regrettably, it is rare for these technologies to be deployed in production because of their hidden costs. For example, model-based approaches for configuration and optimization are powerful in laboratory demonstrations, but these approaches typically have high costs for model construction and maintenance. This talk discusses the need for a methodology for engineering autonomic systems that systematically addresses requirements, design, implementation, and assessment. The approach discussed in this talk is based on control theory, an approach that is widely used in other engineering disciplines to blend formal mathematics with practical insights to build robust systems.