Skip to Main content Skip to Navigation
Theses

Reconciling parallelism expressivity and separation of concerns in reconfiguration of distributed systems

Maverick Chardet 1, 2, 3
1 AVALON - Algorithms and Software Architectures for Distributed and HPC Platforms
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
2 STACK - Software Stack for Massively Geo-Distributed Infrastructures
Inria Rennes – Bretagne Atlantique , LS2N - Laboratoire des Sciences du Numérique de Nantes
Abstract : Distributed computer systems, which run on multiple computers, are now commonplace and used even in critical systems. However, these systems are becoming more and more complex, in terms of scale, dynamicity and expected quality of service. Reconfiguration of distributed systems consists in changing their state at runtime. Distributed systems may be reconfigured for many reasons, including deploying them, updating them, adapting them to fulfill new requirements (in terms of user capacity, energy efficiency, reliability, costs, etc.) or even changing their capabilities. Existing reconfiguration frameworks fall short of providing at the same time parallelism expressivity and separation of concerns between the different actors interacting with the system. The focus of this thesis is to prove that these properties can be reconciled by modelling precisely the life-cycle of each module of distributed systems, while providing appropriate interfaces between the different levels of conception. Two formal models implementing this idea are provided, one for the specific case of deployment and one for reconfiguration. Evaluation is performed on both synthetic and real use-cases and show that these models have a higher level of parallelism expressivity than their counterparts while conserving a good level of separation of concerns.
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03230476
Contributor : Abes Star :  Contact
Submitted on : Friday, October 22, 2021 - 7:48:28 AM
Last modification on : Wednesday, November 3, 2021 - 6:14:56 AM

File

2020IMTA0194_Chardet-Maverick....
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03230476, version 2

Citation

Maverick Chardet. Reconciling parallelism expressivity and separation of concerns in reconfiguration of distributed systems. Distributed, Parallel, and Cluster Computing [cs.DC]. Ecole nationale supérieure Mines-Télécom Atlantique, 2020. English. ⟨NNT : 2020IMTA0194⟩. ⟨tel-03230476v2⟩

Share

Metrics

Record views

49

Files downloads

65