Jean-Louis Pazat - Academia.edu (original) (raw)

Papers by Jean-Louis Pazat

2018 17th International Symposium on Parallel and Distributed Computing (ISPDC)

It is now feasible for consumers to buy inexpensive devices that can be installed at home and acc... more It is now feasible for consumers to buy inexpensive devices that can be installed at home and accessed remotely thanks to an Internet connection. Such a simple "self-hosting" can be an alternative to traditional cloud providers, especially for privacy-conscious users. We discuss how a community of users can pool their devices in order to host microservices-based applications, where each microservice is deployed on a different device. The performance of such an application depends heavily on the computing and network resources that are available and on the placement of each microservice. Finding the placement that minimizes the application response time is an NP-hard problem. We show that, thanks to well known optimization techniques (Particle Swarm Optimization), it is possible to quickly find a service placement resulting in a response time close to the optimal one. Thanks to an emulation platform, we evaluate the robustness of this solution to changes in the Quality of Service under conditions typical of a residential access network.

2018 IEEE International Conference on Cloud Computing Technology and Science (CloudCom)

Infrastructure as a Service (IaaS) clouds are very dynamic with at runtime frequent changes at di... more Infrastructure as a Service (IaaS) clouds are very dynamic with at runtime frequent changes at different levels of the virtual infrastructure. For cloud tenants, this affects the ability of a security monitoring framework to successfully detect attacks. In this paper, we propose SAIDS, a self-adaptable intrusion detection system for IaaS clouds that is able to adapt its components based on dynamic events that occur in a cloud infrastructure. We implemented and experimentally evaluated SAIDS, and show that it is a scalable solution that successfully detects attacks even during the adaptation process while imposing negligible overhead to cloud operations and tenant applications.

Grid applications have been prone to encountering problems such as failures or malicious attacks ... more Grid applications have been prone to encountering problems such as failures or malicious attacks during execution, due to their dis-tributed and large-scale features. The application itself, however, has limited power to address these problems. This paper presents the design, and implementation of an adaptive framework- Dynasa, which strives to handle security problems using adaptive fault-tolerance (i.e., checkpointing and replication) during the execution of appli-cations according to the status of the grid environments.

Wireless communication based applications always need adaptation techniques because the bandwidth... more Wireless communication based applications always need adaptation techniques because the bandwidth and the resources of the network often change during the execution of an application. These applications are characterized by a low bandwidth, small data sets and low-end machines. At the opposite of the spectrum, Grid applications use powerful supercomputers, high bandwidth networks and process large amount of data. However, during an application run, both the bandwidth of the network and the computing resources may also vary. In this paper, we show than despite the difference between these kinds of applications, Grid applications can benefit from adaptation techniques primarily designed for wireless applications.

Tout d'abord, j'adresse toute ma gratitudeà Frédéric Suter età Jens Gusted d'avoir accepté de dir... more Tout d'abord, j'adresse toute ma gratitudeà Frédéric Suter età Jens Gusted d'avoir accepté de diriger cette thèse, et d'avoirété toujours disponibles tout au long de ces travaux. Je remercieégalement, pour l'aide qu'elle nous a apportée, toute l'équipe de maintenance de la bibliothèque logicielle SimGrid utilisée pour la mise en oeuvre de notre simulateur. Il s'agit notamment d'Arnaud Legrand qui a intégré un modèle de tâches parallèles dans SimGridà notre demande, de Martin Quinson et Malek Cherier que j'ai souvent sollicités pour mettreà disposition une version récente de cette bibliothèque. Un grand merci aussià Henri Casanova età Pierre-François Dutot avec qui nous avons collaborer durant cette thèse. Je voudraiségalement remercier les membres de mon jury qui ontété tous présents lors de la soutenance et qui sont venus de loin pour la plupart d'entre eux. Mercià Denis Trystram, pour l'honneur qu'il m'a fait d'être président de ce jury. Mercià Olivier Beaumont et Jean-Louis Pazat d'avoir accepté d'être les rapporteurs de cette thèse et, pour leurs critiques pertinentes. Merci aux examinateurs Marie-Claude Portmann, Thomas Rauber et André Schaff. Je ne saurais oublier tous les autres membres de l'équipe AlGorille avec qui j'ai eu d'excellents rapports, en particulier

Grid architectures are execution environments that are known to be at the same time distributed, ... more Grid architectures are execution environments that are known to be at the same time distributed, parallel, heterogeneous and dynamic. While current tools focus solutions for hiding distribution, parallelism and heterogeneity, this approach does not t well their dynamic aspect. Indeed, if applications are able to adapt themselves to environmental changes, they can benet from it to achieve better performance. This article presents Afpac, a model extending Dynaco for designing self-adaptable parallel components that can be assembled to build applications for Grid. This model includes the denition of a consistency criterion for the dynamic adaptation of SPMD components. We propose a solution to implement this criterion. It has been evalued using both synthetic and real codes to exhibit the behavior of several proposed strategies.

This paper presents the parallelization process of a Wave Propagation application using the Pando... more This paper presents the parallelization process of a Wave Propagation application using the Pandore environment. Tools are briefly described, the stress is put on the description of the parallelization by data distribution and performance results are shown.

2010 10th IEEE International Conference on Computer and Information Technology

Cet article propose une solution pour heberger des services a destination d'une communaute su... more Cet article propose une solution pour heberger des services a destination d'une communaute sur des equipements partages par ses membres. Ces equipements possedent des ressources de calcul, de stockage et de communication heterogenes ainsi qu'une disponibilite non garantie. Afin de s'adapter a ces contraintes, nous proposons une approche qui s'appuie sur les micro-services virtualises dans des conteneurs logiciels. Cette solution a ete mise en oeuvre sur des equipements reels distribues chez des utilisateurs afin de valider la pertinence de l'approche.

Management Summary This deliverable is aimed at summarizing the joint research in WP-JRA-2.3. rel... more Management Summary This deliverable is aimed at summarizing the joint research in WP-JRA-2.3. related to decision support for local adaptation. It is an intermediate stage on the research roadmap, starting from issues of local adaptation and self-healing (CD-JRA-2.3.2) to the most complex case involving distributed multi-level adaptation (CD-JRA-2.3.8), where we investigate and integrate certain methods and techniques incrementally. The work is based on and motivated by the antecedent deliverable ''Basic requirements for self-healing services and decision support for local adaptation'' (CD-JRA-2.3.2) and is focused on local adaptation and decision which we consider one of the most important ways to investigate the applicability of certain policies to trigger local adaptation mechanism, and is organized around the general adaptation framework introduced in CD-JRA-2.3.2. Results are presented in 10 published papers that constitute the core contribution of this delivera...

In this paper, we present the overall design of Pandore II, an Environment dedicated to the exper... more In this paper, we present the overall design of Pandore II, an Environment dedicated to the experimentation of distribution of sequential programs for their execution on distributed memory parallel architectures. The emphasis is then put on two performance analysis tools integrated in this environment.

This paper presents an environment for programming distributed memory computers using HPF-like da... more This paper presents an environment for programming distributed memory computers using HPF-like data distribution features. Emphasis is put on compilation techniques and distributed array management. Results are shown for some well known numerical algorithms.

We consider the problem of executing composite computing applications called workflows on top of ... more We consider the problem of executing composite computing applications called workflows on top of unreliable computing infrastructures. Having in mind the situation of the electric delivery in the sub-saharan area, we propose BEDWE, a decentralized workflow engine able to dynamically assign portions of the workflow to currently live compute nodes. More precisely, in a point-to-point manner, each node can receive a part of the workflow and delegate a subpart of it to another node. This mechanism can be repeated recursively until the whole workflow is executed. BEDWE includes a mechanism to support nodes leaving the network due to power outage. We present a software prototype of BEDWE and its experimentation over the French nation-wide Grid’5000 platform.

... On peut ainsi représenter exactement la structure de certains algorithmes présentant, une top... more ... On peut ainsi représenter exactement la structure de certains algorithmes présentant, une topologie r.lassinne du type anneau logique, arborescence. Le coût et la complexité des circuits d'aiguil-lage limitent encore l'utilisation de telles architectures. ...

... Hiroki Saito, Tokyo Denki University, Japan Vitaly Klyuev, University of Aizu, Japan Satoru O... more ... Hiroki Saito, Tokyo Denki University, Japan Vitaly Klyuev, University of Aizu, Japan Satoru Ohta, Toyama Prefectural University, Japan ... Ryuichi Oka, University of Aizu, Japan Andrew Simmonds, University of Technology, Australia Alexander Vazhenin, University of Aizu, Japan ...

Ce document decrit les methodes de compilation et d'execution pour la generation automatique ... more Ce document decrit les methodes de compilation et d'execution pour la generation automatique de code distribue par distribution de donnees.

2018 17th International Symposium on Parallel and Distributed Computing (ISPDC)

It is now feasible for consumers to buy inexpensive devices that can be installed at home and acc... more It is now feasible for consumers to buy inexpensive devices that can be installed at home and accessed remotely thanks to an Internet connection. Such a simple "self-hosting" can be an alternative to traditional cloud providers, especially for privacy-conscious users. We discuss how a community of users can pool their devices in order to host microservices-based applications, where each microservice is deployed on a different device. The performance of such an application depends heavily on the computing and network resources that are available and on the placement of each microservice. Finding the placement that minimizes the application response time is an NP-hard problem. We show that, thanks to well known optimization techniques (Particle Swarm Optimization), it is possible to quickly find a service placement resulting in a response time close to the optimal one. Thanks to an emulation platform, we evaluate the robustness of this solution to changes in the Quality of Service under conditions typical of a residential access network.

2018 IEEE International Conference on Cloud Computing Technology and Science (CloudCom)

Infrastructure as a Service (IaaS) clouds are very dynamic with at runtime frequent changes at di... more Infrastructure as a Service (IaaS) clouds are very dynamic with at runtime frequent changes at different levels of the virtual infrastructure. For cloud tenants, this affects the ability of a security monitoring framework to successfully detect attacks. In this paper, we propose SAIDS, a self-adaptable intrusion detection system for IaaS clouds that is able to adapt its components based on dynamic events that occur in a cloud infrastructure. We implemented and experimentally evaluated SAIDS, and show that it is a scalable solution that successfully detects attacks even during the adaptation process while imposing negligible overhead to cloud operations and tenant applications.

Grid applications have been prone to encountering problems such as failures or malicious attacks ... more Grid applications have been prone to encountering problems such as failures or malicious attacks during execution, due to their dis-tributed and large-scale features. The application itself, however, has limited power to address these problems. This paper presents the design, and implementation of an adaptive framework- Dynasa, which strives to handle security problems using adaptive fault-tolerance (i.e., checkpointing and replication) during the execution of appli-cations according to the status of the grid environments.

Wireless communication based applications always need adaptation techniques because the bandwidth... more Wireless communication based applications always need adaptation techniques because the bandwidth and the resources of the network often change during the execution of an application. These applications are characterized by a low bandwidth, small data sets and low-end machines. At the opposite of the spectrum, Grid applications use powerful supercomputers, high bandwidth networks and process large amount of data. However, during an application run, both the bandwidth of the network and the computing resources may also vary. In this paper, we show than despite the difference between these kinds of applications, Grid applications can benefit from adaptation techniques primarily designed for wireless applications.

Tout d'abord, j'adresse toute ma gratitudeà Frédéric Suter età Jens Gusted d'avoir accepté de dir... more Tout d'abord, j'adresse toute ma gratitudeà Frédéric Suter età Jens Gusted d'avoir accepté de diriger cette thèse, et d'avoirété toujours disponibles tout au long de ces travaux. Je remercieégalement, pour l'aide qu'elle nous a apportée, toute l'équipe de maintenance de la bibliothèque logicielle SimGrid utilisée pour la mise en oeuvre de notre simulateur. Il s'agit notamment d'Arnaud Legrand qui a intégré un modèle de tâches parallèles dans SimGridà notre demande, de Martin Quinson et Malek Cherier que j'ai souvent sollicités pour mettreà disposition une version récente de cette bibliothèque. Un grand merci aussià Henri Casanova età Pierre-François Dutot avec qui nous avons collaborer durant cette thèse. Je voudraiségalement remercier les membres de mon jury qui ontété tous présents lors de la soutenance et qui sont venus de loin pour la plupart d'entre eux. Mercià Denis Trystram, pour l'honneur qu'il m'a fait d'être président de ce jury. Mercià Olivier Beaumont et Jean-Louis Pazat d'avoir accepté d'être les rapporteurs de cette thèse et, pour leurs critiques pertinentes. Merci aux examinateurs Marie-Claude Portmann, Thomas Rauber et André Schaff. Je ne saurais oublier tous les autres membres de l'équipe AlGorille avec qui j'ai eu d'excellents rapports, en particulier

Grid architectures are execution environments that are known to be at the same time distributed, ... more Grid architectures are execution environments that are known to be at the same time distributed, parallel, heterogeneous and dynamic. While current tools focus solutions for hiding distribution, parallelism and heterogeneity, this approach does not t well their dynamic aspect. Indeed, if applications are able to adapt themselves to environmental changes, they can benet from it to achieve better performance. This article presents Afpac, a model extending Dynaco for designing self-adaptable parallel components that can be assembled to build applications for Grid. This model includes the denition of a consistency criterion for the dynamic adaptation of SPMD components. We propose a solution to implement this criterion. It has been evalued using both synthetic and real codes to exhibit the behavior of several proposed strategies.

This paper presents the parallelization process of a Wave Propagation application using the Pando... more This paper presents the parallelization process of a Wave Propagation application using the Pandore environment. Tools are briefly described, the stress is put on the description of the parallelization by data distribution and performance results are shown.

2010 10th IEEE International Conference on Computer and Information Technology

Cet article propose une solution pour heberger des services a destination d'une communaute su... more Cet article propose une solution pour heberger des services a destination d'une communaute sur des equipements partages par ses membres. Ces equipements possedent des ressources de calcul, de stockage et de communication heterogenes ainsi qu'une disponibilite non garantie. Afin de s'adapter a ces contraintes, nous proposons une approche qui s'appuie sur les micro-services virtualises dans des conteneurs logiciels. Cette solution a ete mise en oeuvre sur des equipements reels distribues chez des utilisateurs afin de valider la pertinence de l'approche.

Management Summary This deliverable is aimed at summarizing the joint research in WP-JRA-2.3. rel... more Management Summary This deliverable is aimed at summarizing the joint research in WP-JRA-2.3. related to decision support for local adaptation. It is an intermediate stage on the research roadmap, starting from issues of local adaptation and self-healing (CD-JRA-2.3.2) to the most complex case involving distributed multi-level adaptation (CD-JRA-2.3.8), where we investigate and integrate certain methods and techniques incrementally. The work is based on and motivated by the antecedent deliverable ''Basic requirements for self-healing services and decision support for local adaptation'' (CD-JRA-2.3.2) and is focused on local adaptation and decision which we consider one of the most important ways to investigate the applicability of certain policies to trigger local adaptation mechanism, and is organized around the general adaptation framework introduced in CD-JRA-2.3.2. Results are presented in 10 published papers that constitute the core contribution of this delivera...

In this paper, we present the overall design of Pandore II, an Environment dedicated to the exper... more In this paper, we present the overall design of Pandore II, an Environment dedicated to the experimentation of distribution of sequential programs for their execution on distributed memory parallel architectures. The emphasis is then put on two performance analysis tools integrated in this environment.

This paper presents an environment for programming distributed memory computers using HPF-like da... more This paper presents an environment for programming distributed memory computers using HPF-like data distribution features. Emphasis is put on compilation techniques and distributed array management. Results are shown for some well known numerical algorithms.

We consider the problem of executing composite computing applications called workflows on top of ... more We consider the problem of executing composite computing applications called workflows on top of unreliable computing infrastructures. Having in mind the situation of the electric delivery in the sub-saharan area, we propose BEDWE, a decentralized workflow engine able to dynamically assign portions of the workflow to currently live compute nodes. More precisely, in a point-to-point manner, each node can receive a part of the workflow and delegate a subpart of it to another node. This mechanism can be repeated recursively until the whole workflow is executed. BEDWE includes a mechanism to support nodes leaving the network due to power outage. We present a software prototype of BEDWE and its experimentation over the French nation-wide Grid’5000 platform.

... On peut ainsi représenter exactement la structure de certains algorithmes présentant, une top... more ... On peut ainsi représenter exactement la structure de certains algorithmes présentant, une topologie r.lassinne du type anneau logique, arborescence. Le coût et la complexité des circuits d'aiguil-lage limitent encore l'utilisation de telles architectures. ...

... Hiroki Saito, Tokyo Denki University, Japan Vitaly Klyuev, University of Aizu, Japan Satoru O... more ... Hiroki Saito, Tokyo Denki University, Japan Vitaly Klyuev, University of Aizu, Japan Satoru Ohta, Toyama Prefectural University, Japan ... Ryuichi Oka, University of Aizu, Japan Andrew Simmonds, University of Technology, Australia Alexander Vazhenin, University of Aizu, Japan ...

Ce document decrit les methodes de compilation et d'execution pour la generation automatique ... more Ce document decrit les methodes de compilation et d'execution pour la generation automatique de code distribue par distribution de donnees.