Title :
On the quorum requirement and value selection rule for Fast Paxos
Author_Institution :
Dept. of Electr. & Comput. Eng., Cleveland State Univ., Cleveland, OH, USA
Abstract :
Fault tolerance is a crucial requirement for big data and cloud computing because it helps remove single point of failures. The Paxos algorithm has been employed in a number of real-world fault tolerant systems. Fast Paxos is one of the latest algorithms in the Paxos family of algorithms for distributed consensus. In Fast Paxos, an acceptor is allowed to cast a vote for a value of its own choice unilaterally in a fast round, which eliminates a communication step for reaching consensus. As a tradeoff, a bigger quorum than simple majority must be used and a value selection rule is employed to select the right value for agreement to overcome collisions. The derivation of quorum requirements and value selection rule in the original article on Fast Paxos by Lamport, while very rigorous, may be hard to follow and difficult to implement. In this paper, we present a more intuitive approach to deriving the quorum requirements and easy-to-implement value selection rule for Fast Paxos.
Keywords :
Big Data; cloud computing; fault tolerant computing; Big Data; cloud computing; fast Paxos algorithm; fault tolerant systems; quorum requirement; value selection rule; Big data; Cloud computing; Fault tolerance; Fault tolerant systems; Proposals; Servers; Cloud Computing; Distributed Consensus; Fault Tolerance; Quorum Requirements; State-Machine Replication;
Conference_Titel :
Software Engineering and Service Science (ICSESS), 2014 5th IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4799-3278-8
DOI :
10.1109/ICSESS.2014.6933593