The Design of a Group Communication System for Distributed and Complex Systems

Distributed and Complex Systems are more and more vulnerable to failures on account of their growing complexity and distribution. This problem motivates the need for a fault tolerance mechanism in such systems. Hence, we focus on two reviews: (1) the distributed algorithmic review, and (2) the development of the component systems that make the task of failure detection and repair much easier.

Our paper highlights the first one approach. Indeed, it focuses on the problem of reliable duplication mechanism on grid environments.

As a first task, we define a new group communication system extending the FSR protocol which is developed by the INRIA/EPFL. Because of his interesting results in a cluster of homogenous machines, this work consists on adapting this system to the grid environments. Whereas, the following part of our paper, shows architecture of a fault-tolerant distributed system. So in order to reach our objective and achieve our research, this is our planning:

• A group communication system algorithmic review. 
• What is a group communication system? 
• The group communication system FSR([ 2]) Introducing our group communication system for grid environment denoted FSRG. 
• The simulation on JSim to evaluate the performance of FSRG.