Requirements for high performance group support in distributed systems
Werner Vogels, Paulo Verı́ssimo, Luı́s Rodrigues · 1992
Although the use of groups and group communication is becoming more and more accepted as an important tool for implementing distributed applications and algorithms, there is still much controversy about the way the group paradigm should actually be implemented. In this paper the authors identify some of the areas in which these controversies still exist and express their views on the matters causing the disagreements. Introduction. The group paradigm is widely accepted as being an excellent method of structuring activities in distributed systems. From the pioneering projects in the past [1,7,5], a large number of research projects in areas related with group structuring and reliable group communication have emerged [2,8,9,10,13,14]. Although there is consensus on the merits of the group paradigm, there is hardly any of this agreement on how the paradigm should actually be implemented and presented to the user. Recently, especially in Europe, we have noticed how far the ideas on relate...