Cheap Paxos

  title={Cheap Paxos},
  author={Leslie Lamport and Mike Massa},
  journal={International Conference on Dependable Systems and Networks, 2004},
  • L. Lamport, M. Massa
  • Published 28 June 2004
  • Computer Science
  • International Conference on Dependable Systems and Networks, 2004
Asynchronous algorithms for implementing a fault-tolerant distributed system, which can make progress despite the failure of any F processors, require 2F + 1 processors. Cheap Paxos, a variant of the Paxos algorithm, guarantees liveness under the additional assumption that the set of nonfaulty processors does not "jump around" too fast, but uses only F + 1 main processors that actually execute the system and F auxiliary processors that are used only to handle the failure of a main processor… 

Unbounded Pipelining in Dynamically Reconfigurable Paxos Clusters

This article shows how to extend Paxos to support a more general form of reconfiguration which subsumes the original bounded-pipeline approach as well as Raft-like fullyconcurrent reconfigurations and more besides.

CheapBFT: resource-efficient byzantine fault tolerance

CheapBFT is presented, a BFT system that tolerates that all but one of the replicas active in normal-case operation become faulty, and which allows the system to safely switch to another, more resilient agreement protocol.


  • Computer Science
  • 2021
Donut Paxos and Donut MultiPaxos are presented, a reconfigurable consensus and state machine replication protocol respectively, and they present a framework that can be generalized to other replication protocols in a way that previous reconfiguration techniques can not.

Cheap and Available State Machine Replication

The key contribution of this paper is to identify that, when agreement nodes and execution nodes are logically separated, they each presents a unique property that enables lazy recovery, which is applied to Paxos and built ThriftyPaxos, which, as shown in the evaluation, can achieve higher throughput and similar availability comparing to standard Paxos.

Time-efficient asynchronous service replication

This thesis investigates the cost associated with implementing fundamental abstractions constituting the core of service replication in asynchronous distributed systems, namely (a) consensus and (b) the read/write register, and determines how quickly consensus can be solved in the common case.

Increasing performance in byzantine fault-tolerant systems with on-demand replica consistency

An extension to existing BFT architectures is proposed that increases performance for the default number of replicas by optimizing the resource utilization of their execution stages, and shows that with each replica executing only a part of all requests, the overall performance of a Byzantine fault-tolerant NFS can be almost doubled.

Reconfigurable Atomic Transaction Commit

This work codifies the core ideas of FARM as distributed TCS protocols, rigorously proves them correct and highlights the trade-offs required by the use of RDMA.

Formal Specification and Safety Proof of a Leaderless Concurrent Atomic Broadcast Algorithm

This work provides a formal specification of AllConcur that enables a better understanding of the algorithm, and formally proves AllConCur's safety property on the basis of this specification.

Ring Paxos: A high-throughput atomic broadcast protocol

Ring Paxos is a protocol derived from Paxos that inherits the reliability of Paxos and can be implemented very efficiently, and is compared to other atomic broadcast protocols.

Matchmaker Paxos: A Reconfigurable Consensus Protocol [Technical Report]

Matchmaker Paxos and Matchmaker MultiPaxos are presented, a reconfigurable consensus and state machine replication protocol respectively, and they present a framework that can be generalized to other replication protocols in a way that previous reconfiguration techniques can not.



How to Build a Highly Available System Using Consensus

The general scheme for efficient highly available computing is explained, a general method for understanding concurrent and fault-tolerant programs is given, and the Paxos algorithm is derived as an example of the method.

Revisiting the PAXOS algorithm

Consensus in the presence of partial synchrony

Fault-tolerant consensus protocols are given for various cases of partial synchrony and various fault models that allow partially synchronous processors to reach some approximately common notion of time.

Time, clocks, and the ordering of events in a distributed system

A distributed algorithm is given for synchronizing a system of logical clocks which can be used to totally order the events, and a bound is derived on how far out of synchrony the clocks can become.

Implementing fault-tolerant services using the state machine approach: a tutorial

The state machine approach is a general method for implementing fault-tolerant services in distributed systems and protocols for two different failure models—Byzantine and fail stop are described.

Replication in the harp file system

Preliminary performance results indicate that Harp provides equal or better response time and system capacity than an unreplicated implementation of NFS that uses Unix files directly.

Voting with regenerable volatile witnesses

It is shown that under standard Markovian assumptions two full replicas and one regenerable volatile witness managed by a two-tier dynamic voting protocol provide a higher data availability than three full Replicas managed by majority consensus voting or optimistic dynamic voting provided site failures can be detected significantly faster than they can be repaired.

The part-time parliament

The Paxon parliament's protocol provides a new way of implementing the state machine approach to the design of distributed systems.

Paxos Made Simple

The Paxos algorithm, when presented in plain English, is very simple and straightforward to understand.

The Implementation of Reliable Distributed Multiprocess Systems