Sites Inria

Version française

DELYS Research team

DistributEd aLgorithms and sYStems

Team presentation

DELYS addresses both theoretical and practical issues of distributed systems. The research covers a large spectrum of distributed computer systems including cloud and edge computing systems, mobile networks, dynamic communicating entities such network of robots, multicore computers...

These new distributed computing systems are subject to many important features which include scalability, fault tolerance, dynamics, heterogeneity, and virtualisation at many levels. Algorithms designed for traditional static distributed systems need to be redefined to consider the unstable nature of the distributed system.

 

Research themes

DELYS will focus on three key challenges:

  1. Rethinking distributed algorithms. From a theoretical point of view the key question is how to adapt the fundamental building blocks to new dynamic architectures. DELYS’s objectives are then (1) to identify under which realistic assumptions a given fundamental problem such as mutual exclusion, consensus or leader election can be solved and (2) to design efficient algorithms under these assumptions.
  2. Resources management in heterogeneous systems. Managing resources in such systems requires fully decentralized solutions. In this context, data management is a key component. The fundamental issue we address in DELYS is how to efficiently and reliably share information in highly distributed environments.
  3. Adaptation of runtimes. With the increasingly widespread use of multicore architectures and virtualized environments, internal runtime protocols need to be revisited. Our research will thus focus on providing solutions to efficiently share resources such as memory between virtualized environments without jeopardizing isolation properties.

International and industrial relations

Academic collaborations (industrial / international)

  • RainbowFS ANR
  • Wolphin FUI
  • LightKone –H2020 project
  • PaDMetBio – Stic Amsud
  • CHOOSING – Capes-Cofecub 

Billateral industrial collaborations

  • Scality SA, Paris
  • Orange, France

Keywords: Distibuted algorithms and systems Data replication and consistency Fault Tolerance Large scale systems Virtual Machine and Multi-core Dynamic networks