| | |
Summary: Tight Bounds for Asynchronous Randomized Consensus
(Preliminary Version)
Hagit Attiya
Department of Computer Science
Technion
Haifa, Israel
hagit@cs.technion.ac.il
Keren Censor
Department of Computer Science
Technion
Haifa, Israel
ckeren@cs.technion.ac.il
ABSTRACT
A distributed consensus algorithm allows n processes to
reach a common decision value starting from individual in-
puts. Wait-free consensus, in which a process always termi-
nates within a finite number of its own steps, is impossible
in an asynchronous shared-memory system. However, con-
sensus becomes solvable using randomization when a pro-
cess only has to terminate with probability 1. Randomized
|
