I am the leader of the VeRDi project is funded by the French region Pays De La Loire where Nantes is located. VeRDi is an acronym for Verified Reconfiguration Driven by execution. It aims at addressing distributed software reconfiguration in an efficient and verified way.
I am one of the members of the Lowcomote European project. Lowcomote aims to train a generation of professionals in the design, development and operation of new LCDPs, that overcome the limitations above, by being scalable (i.e., supporting the development of large-scale applications, and using artefacts coming from a large number of users), open (i.e., based on interoperable and exchangeable programming models and standards), and heterogeneous (i.e., able to integrate with models coming from different engineering disciplines). These scientists will drive the upgrade of the current landscape of Low-Code Development Platforms to Low-Code Engineering Platforms (LCEPs).
I am involved in the HYDDA project aims to develop a software solution allowing the deployment of Big Data applications (with hybrid design (HPC/CLoud)) on heterogeneous platforms (cluster, Grid, private Cloud) and the orchestration of computation tasks (like Slurm, Nova for OpenStack, or Swarm for Docker). The mains challenges addressed by the project are:
- How to propose an easy-to-use service to host application components (from deployment to supression) that are both typed Cloud and HPC?
- How to propose a service that unifies the HPCaaS (HPC as a service) and the Infrastructure as a Service (IaaS) in order to offer on-demand resources and to take into account specificities of scientific applications?
- How optimize resources usage of these platforms (CPU, RAM, Disk, Energy, etc.) in order to propose solutions at the least cost?
Former projects: The Discovery inria project lab
- Co-chair of the working groupe YODA (trustworthY and Optimal Dynamic Adaptation) in the national GDR GPL (software engineering and languages)
- Publicity co-chair of CCGrid 2021
- Track vice-chair of CCGrid 2020, track “Programming models and runtime systems”
- Track vice-chair of the IEEE BigData Congress 2018, the track “Quality of Big Data Services”
- PC member of the French working group GDR GPL (CNRS) 2020 new challenges
- PC member of Euro-par 2020
- PC member of ICCS 2020 (since 2015)
- PC member of ISCC 2020
- PC member CCGrid 2019
- PC member ClouCom 2018
- PC member of CIoT 2018
- PC member of ClouCom 2017
- PC member of ICFEC 2017
- PC member of SAC PAPP (since 2016)
- PC member of Compas (since 2015)
- PC member of the poster sessino of CCGrid’16
Permanent position juries
- 2020 Selection of an associate professor in Rennes
- 2020 Selection of an associate professor in Sophia Antipolis
PhD and Master juries
- 2019-09-23 Alexandre Da Silva Veith, Université de Lyon, PhD defense, Quality of Service Aware Mechanisms for (Re)Configuring Data Stream Processing Applications on Highly Distributed Infrastructure
- 2019-05-02 Joan Philippe, Northern Arizona University, Master thesis, systematic development of efficient programs on parallel data structures
- 2018-12-21 Stéphanie Chillita, Université de Lille, PhD defense, Inferring Models from Cloud APIs and Reasoning over Them: A Tooled and Formal Approach
- 2018-06-05 Gustavo Sousa, Université de Lille, PhD defense, A Software Product Lines-BasedApproach for the Setup andAdaptation of Multi-CloudEnvironments
Other paper reviews
- Journals: JIoT, ANTE, FGCS, JPDC, IJPP, ParCo, HLPP.
- Review for IPDPS 2018
Concerto is a preliminary implementation in Python 3 of the Concerto reconfiguration model.
- Source code (GPL Licence v3): https://gitlab.inria.fr/mchardet/madpp
Madeus Application Deployer (MAD)
MAD is a PYTHON implementation of the Madeus model. Its purpose is to offer a way to define a low-level deployment process and coordinate this deployment in an efficient and reliable manner. MAD is low-level but offers a highly generic deployment model.
- Documentation: https://mad.readthedocs.io/en/latest/
- Source code (GPL Licence v3): https://gitlab.inria.fr/Madeus/mad
The SkelGIS Library
The SkelGIS Library has been developed during my Ph.D. SkelGIS is a C++-embedded domain specific language for numerical simulations based on explicit schemes. SkelGIS has been implemented for 2D Cartesian meshes and for network of Cartesian meshes. It also has been evaluated on two real case numerical simulations. I am for now the only contributor of the Library, however it is no longer maintained. You can download sources here.