Toward Run-time Coordination of Reconfiguration ...
Document type :
Communication dans un congrès avec actes
Title :
Toward Run-time Coordination of Reconfiguration Requests in Cloud Computing Systems
Author(s) :
Farhat, Salman [Auteur]
Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Bliudze, Simon [Auteur]
Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Duchien, Laurence [Auteur]
Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Kouchnarenko, Olga [Auteur]
Franche-Comté Électronique Mécanique, Thermique et Optique - Sciences et Technologies (UMR 6174) [FEMTO-ST]
Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Bliudze, Simon [Auteur]

Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Duchien, Laurence [Auteur]
Self-adaptation for distributed services and large software systems [SPIRALS]
Université de Lille
Kouchnarenko, Olga [Auteur]
Franche-Comté Électronique Mécanique, Thermique et Optique - Sciences et Technologies (UMR 6174) [FEMTO-ST]
Scientific editor(s) :
Sung-Shik Jongmans
Conference title :
COORDINATION 2023 - 25th International Conference on Coordination Models and Languages
City :
Lisbon
Country :
Portugal
Start date of the conference :
2023-06-19
Publisher :
Springer Nature Switzerland
Publication date :
2023-04-22
English keyword(s) :
Concurrent Component-based Systems
Variability Models
Self-Configuration
Dynamic Reconfiguration
Variability Models
Self-Configuration
Dynamic Reconfiguration
HAL domain(s) :
Informatique [cs]/Génie logiciel [cs.SE]
English abstract : [en]
Cloud applications and cyber-physical systems are becoming increasingly complex, requiring frequent reconfiguration to adapt to changing needs and requirements. Existing approaches compute new valid configurations either ...
Show more >Cloud applications and cyber-physical systems are becoming increasingly complex, requiring frequent reconfiguration to adapt to changing needs and requirements. Existing approaches compute new valid configurations either at design time, at runtime, or both. However, these approaches can lead to significant computational or validation overheads for each reconfiguration step. We propose a component-based approach that avoids computational and validation overheads using a representation of the set of valid configurations as a variability model. More precisely, our approach leverages feature models to automatically generate, in a component-based formalism called JavaBIP, run-time variability models that respect the feature model constraints. Produced run-time variability models enable control over application reconfiguration by executing reconfiguration requests in such a manner as to ensure the (partial) validity of all reachable configurations. We evaluate our approach on a simple web application deployed on the Heroku cloud platform. Experimental results show that the overheads induced by generated run-time models on systems involving up to 300 features are negligible, demonstrating the practical interest of our approach.Show less >
Show more >Cloud applications and cyber-physical systems are becoming increasingly complex, requiring frequent reconfiguration to adapt to changing needs and requirements. Existing approaches compute new valid configurations either at design time, at runtime, or both. However, these approaches can lead to significant computational or validation overheads for each reconfiguration step. We propose a component-based approach that avoids computational and validation overheads using a representation of the set of valid configurations as a variability model. More precisely, our approach leverages feature models to automatically generate, in a component-based formalism called JavaBIP, run-time variability models that respect the feature model constraints. Produced run-time variability models enable control over application reconfiguration by executing reconfiguration requests in such a manner as to ensure the (partial) validity of all reachable configurations. We evaluate our approach on a simple web application deployed on the Heroku cloud platform. Experimental results show that the overheads induced by generated run-time models on systems involving up to 300 features are negligible, demonstrating the practical interest of our approach.Show less >
Language :
Anglais
Peer reviewed article :
Oui
Audience :
Internationale
Popular science :
Non
ANR Project :
Comment :
Part 6: Run-Time Changes
Collections :
Source :
Files
- document
- Open access
- Access the document
- Coordination-23.pdf
- Open access
- Access the document
- document
- Open access
- Access the document
- Coordination-23.pdf
- Open access
- Access the document