Sites Inria

Version française

ASAP Research team

As Scalable As Possible: foundations of large scale dynamic distributed systems

  • Leader : François Taiani
  • Research center(s) : CRI Rennes - Bretagne Atlantique
  • Field : Networks, Systems and Services, Distributed Computing
  • Theme : Distributed Systems and middleware
  • Partner(s) : Université Rennes 1,Institut national des sciences appliquées de Rennes

Team presentation

Our objective is to provide a set of abstractions and algorithms to build serverless large-scale distributed applications involving a large set of volatile, geographically distant, potentially mobile and/or resource limited computing entities.

Epi-Nets Associated team. We have a strong collaboration with Vrije Universiteit ( Prof. Maarten van Steen ) in Amsterdam, NL in the context of the Epi-Nets associated team. Epi-Net addresses several applications using epidemic-based unstructured networks.

Research themes

Our research activities range from theoretical bounds to practical protocols and implementations for large-scale distributed dynamic systems to cope with the recent and tremendous evolution of distributed systems. Effectively we observed huge evolutions.
  • Scale shift in terms of system size, geographical spread, volume of data
  • Dynamic behavior due to versatility, mobility, connectivity patterns
These characteristics lead to a large amount of uncertainty. Mastering such uncertainty is actually our goal.

Applications: We aim at providing a wide range of applications (from content delivery networks to sensors networks, from backup systems to voice over IP, from publish-subscribe systems to genomic databases). We focus our research on two main areas: information management and dissemination. We believe such services are basic building blocks of many distributed applications in two networking contexts: Internet and wireless sensors. These two classes of applications, although exhibiting very different behaviors and constraints, clearly require scalable solutions.

To achieve this ambitious goal, we need to tackle the issues both along the theoretical and practical sides of scalable distributed computing and ASAP will be organized along the following themes:
  • Models and abstractions: dealing with dynamics Although this research will be led by Achour Mostefaoui and Michel Raynal, Marin Bertier and Anne-Marie Kermarrec will also contribute to some specific areas in order to carry on collaborations recently started. Also the outcome of this research theme will be used as inputs to the other research activities. Research topics in this theme include: distributed computing models, distributed computability,and distributed computing abstractions
  • Resource management: dealing with Internet-scale. Managing resources on a large-scale whether they are computing resources, data, events, bandwidth, requires some fully decentralized solutions. Our research in this area focuses on building the relevant overlay networks to provide core functionalities of resource management and discovery. This research will be led by Anne-Marie Kermarrec, Fabrice Le Fessant, Erwan Le Merrer, and Etienne Rivière.
  • Data collection and propagation in sensor networks: In this area, recently initiated, we are investigating the use of P2P algorithms in sensor network systems. At the moment, this research area is essentially studied by the network community although many of these problematics are similar to distributed computing ones like information propagation, resource discovery, etc. This research will be led by Marin Bertier, Yann Busnel, Anne-Marie Kermarrec, Erwan le Merrer, Aline Carneiro Viana.

International and industrial relations

  • Epi-Net associated team: "Building unstructured peer to peer overlays using epidemics", with Vrije Universiteit, Amsterdam, NL
  • Academic projects and collaborations:
    • ReSIST project: the Network of Excellence (NoE), "Towards a global dependability and security framework"
    • LAFMI project: the Franco-Mexican Research Laboratory in Computer Science, with Professor Sergio Rajsbaum
    • Hong Kong Polytechnic University, Hong Kong: with Professor Jiannong Cao
    • Univerity of Madrid, Spain: with Professor Antonio Fernandez
    • University of California, Santa Barbara: with Amr El Abaddi and Divyakant Agrawal
    • ANR MD (Masse de données) Alpage project
    • RNRT SVP project
    • ARC INRIA Recall project: with INRIA project-team CASSIS and REGAL, LIRMM and EPFL
  • Industrial collaborations: with France Telecom R&D, Lannion, on applying P2P techniques to telecom operator frameworks, and with Advestigo on providing expertise in the area of monitoring P2P systems.