Abstract
Abstract: The deployment of a common platform of group services is considered the base for structuring fault tolerant distributed applications. Although a number of solutions are available in literature, the identification of the protocol that better fits the given application requirements is still hard because of the lack of a unique problem statement and of a clear definition of the achievable quality of service. This paper provides a common framework of analysis, identifies the critical design issues that can affect the quality of service, and uses simulations to evaluate the role they play in coping with different failure conditions.