## Mutual Exclusion in Asynchronous Systems with Failure Detectors (2002)

Venue: | J. Parallel Distrib. Comput |

Citations: | 16 - 3 self |

### BibTeX

@ARTICLE{Delporte-Gallet02mutualexclusion,

author = {Carole Delporte-Gallet and Hugues Fauconnier and Rachid Guerraoui and Petr Kouznetsov},

title = {Mutual Exclusion in Asynchronous Systems with Failure Detectors},

journal = {J. Parallel Distrib. Comput},

year = {2002},

volume = {65},

pages = {2005}

}

### Years of Citing Articles

### OpenURL

### Abstract

This paper defines the fault-tolerant mutual exclusion problem in a message-passing asynchronous system and determines the weakest failure detector to solve the problem. This failure detector, which we call the trusting failure detector, and which we denote by , is strictly weaker than the perfect failure detector but strictly stronger than the eventually perfect failure detector #P. The paper shows that a majority of correct processes is necessary to solve the problem with .Moreover,T is also the weakest failure detector to solve the fault-tolerant group mutual exclusion problem.

### Citations

1550 |
Distributed Algorithms
- Lynch
- 1996
(Show Context)
Citation Context ... D is reducible to D ′ in E. 3 The fault-tolerant mutual exclusion problem We define here the fault-tolerant mutual exclusion problem (from now on - FTME) using the terminology and notations given i=-=n [14]. We a-=-ssociate to every process i ∈ Πauser ui that can require exclusive access to the critical section. The users can be thought of as application programs. As in [14], every process i ∈ Π and every ... |

1438 | M.S.: Impossibility of distributed consensus with one faulty process
- Fischer, Lynch, et al.
- 1985
(Show Context)
Citation Context ...ly. In this sense, the problem is related to the famous impossibility result that consensus cannot be solved deterministically in an asynchronous system that is subject to even a single crash failure =-=[6]-=-. To circumvent the impossibility of consensus, Chandra and Toueg [3] introduced the notion of failure detector. Informally, a failure detector is a distributed oracle that gives (possibly incorrect) ... |

903 | Unreliable failure detectors for reliable distributed systems
- Chandra, Toueg
- 1996
(Show Context)
Citation Context ... result that consensus cannot be solved deterministically in an asynchronous system that is subject to even a single crash failure [6]. To circumvent the impossibility of consensus, Chandra and Toueg =-=[3]-=- introduced the notion of failure detector. Informally, a failure detector is a distributed oracle that gives (possibly incorrect) hints about which processes have crashed so far. Each process has acc... |

402 | S.: The weakest failure detector for solving consensus
- Chandra, Hadzilacos, et al.
- 1996
(Show Context)
Citation Context ...rather weak failure detector ✸W is sufficient to solve consensus in an asynchronous system with a majority of correct processes, and that ✸W can be implemented using partial synchrony assumptions.=-= In [2], it-=- is shown that ✸W is also necessary to solve consensus. In short, ✸W is the weakest failure detector to solve consensus. A natural question follows: what is the weakest failure detector to solve t... |

288 |
Solution of a problem in concurrent programming control
- Dijkstra
- 1965
(Show Context)
Citation Context ...n This paper addresses the fault-tolerant mutual exclusion problem in a distributed message-passing system where channels are reliable and processes can fail by crashing. The mutual exclusion problem =-=[5, 11, 12, 17]-=- involves managing access to a single, indivisible resource that can only support one user at atime(mutual exclusion property). The user accessing the resource is said to be in its critical section (C... |

218 |
A sqrt(n) Algorithm for Mutual Exclusion in Decentralized Systems
- Maekawa
- 1985
(Show Context)
Citation Context ... Figure 4 can be substituted with: 9 Cost of resilience wait until receive [ack] fromallj/∈Si. In this section we compare the performance of our algorithm (Figure 3) with the well-known algorithms o=-=f [15]-=- and [18]. 6 The performance of mutual exclusion algorithms can be measured through the following metrics [19]: (a) the bootstrapping delay, which is the time required for a new user before entering t... |

215 |
A new solution of dijkstra’s concurrent programming problem
- Lamport
- 1974
(Show Context)
Citation Context ...n This paper addresses the fault-tolerant mutual exclusion problem in a distributed message-passing system where channels are reliable and processes can fail by crashing. The mutual exclusion problem =-=[5, 11, 12, 17]-=- involves managing access to a single, indivisible resource that can only support one user at atime(mutual exclusion property). The user accessing the resource is said to be in its critical section (C... |

158 |
An optimal algorithm for mutual exclusion in computer networks
- Ricart, Agrawala
- 1981
(Show Context)
Citation Context ... can be substituted with: 9 Cost of resilience wait until receive [ack] fromallj/∈Si. In this section we compare the performance of our algorithm (Figure 3) with the well-known algorithms of [15] an=-=d [18]-=-. 6 The performance of mutual exclusion algorithms can be measured through the following metrics [19]: (a) the bootstrapping delay, which is the time required for a new user before entering the CS for... |

118 |
Algorithms for Mutual Exclusion
- Raynal
- 1986
(Show Context)
Citation Context ...n This paper addresses the fault-tolerant mutual exclusion problem in a distributed message-passing system where channels are reliable and processes can fail by crashing. The mutual exclusion problem =-=[5, 11, 12, 17]-=- involves managing access to a single, indivisible resource that can only support one user at atime(mutual exclusion property). The user accessing the resource is said to be in its critical section (C... |

77 | Abbadi, “An efficient and fault-tolerant solution for distributed mutual exclusion
- Agrawal, El
- 1991
(Show Context)
Citation Context ...lly, mutual exclusion algorithms either consider a failure-free model [13, 19], or suppose that (1) every crash is eventually detected by every correct process and (2) no correct process is suspected =-=[1, 16]-=-: the conjunction of (1) and (2) is equivalent to the assumption of the perfect failure detector P [3]. In other words, perfect information about failures is sufficient to solve the fault-tolerant mut... |

36 |
A Taxonomy of Distributed Mutual Exclusion
- Singhal
- 1993
(Show Context)
Citation Context ... A natural question follows: what is the weakest failure detector to solve the fault-tolerant mutual exclusion problem? Traditionally, mutual exclusion algorithms either consider a failure-free model =-=[13, 19]-=-, or suppose that (1) every crash is eventually detected by every correct process and (2) no correct process is suspected [1, 16]: the conjunction of (1) and (2) is equivalent to the assumption of the... |

28 | Backoff protocols for distributed mutual exclusion and ordering - Chockler, Malkhi, et al. - 2001 |

26 | Asynchronous group mutual exclusion
- Joung
(Show Context)
Citation Context ...ilure detector of [3]) which is strictly weaker than P and which is sufficient to solve the problem even with an arbitrary number of failures. Finally, we turn our attention to group mutual exclusion =-=[8, 9, 10]-=-, a recent generalization of mutual exclusion and we show that T is the weakest to solve fault-tolerant group mutual exclusion (with a majority of correct processes). In other words, we show that the ... |

26 | A simple local spin group mutual exclusion algorithm
- Kean, Moir
- 1999
(Show Context)
Citation Context ...ilure detector of [3]) which is strictly weaker than P and which is sufficient to solve the problem even with an arbitrary number of failures. Finally, we turn our attention to group mutual exclusion =-=[8, 9, 10]-=-, a recent generalization of mutual exclusion and we show that T is the weakest to solve fault-tolerant group mutual exclusion (with a majority of correct processes). In other words, we show that the ... |

14 |
A note on group mutual exclusion
- Hadzilacos
(Show Context)
Citation Context ...ilure detector of [3]) which is strictly weaker than P and which is sufficient to solve the problem even with an arbitrary number of failures. Finally, we turn our attention to group mutual exclusion =-=[8, 9, 10]-=-, a recent generalization of mutual exclusion and we show that T is the weakest to solve fault-tolerant group mutual exclusion (with a majority of correct processes). In other words, we show that the ... |

11 | A Fair Distributed Mutual Exclusion Algorithm
- Lodha, Kshemkalyani
(Show Context)
Citation Context ... A natural question follows: what is the weakest failure detector to solve the fault-tolerant mutual exclusion problem? Traditionally, mutual exclusion algorithms either consider a failure-free model =-=[13, 19]-=-, or suppose that (1) every crash is eventually detected by every correct process and (2) no correct process is suspected [1, 16]: the conjunction of (1) and (2) is equivalent to the assumption of the... |

9 | Analysis of timing-based mutual exclusion with random times
- Gafni, Mitzenmacher
(Show Context)
Citation Context ...ect process in its CS (starvation-freedom property). Besides, we show that if at least one process can crash, then no algorithm can 1 We do not consider here probabilistic mutual exclusion algorithms =-=[4, 7]-=-. 2ssolve the problem using a strictly weaker failure detector. Intuitively, this stems from the fact that, if a process in its CS does not deliberately resign, another process can enter its CS only i... |

5 | An efficient fault-tolerant mutual exclusion algorithm for distributed systems
- Manivannan, Singhal
- 1994
(Show Context)
Citation Context ...lly, mutual exclusion algorithms either consider a failure-free model [13, 19], or suppose that (1) every crash is eventually detected by every correct process and (2) no correct process is suspected =-=[1, 16]-=-: the conjunction of (1) and (2) is equivalent to the assumption of the perfect failure detector P [3]. In other words, perfect information about failures is sufficient to solve the fault-tolerant mut... |

1 |
The mutual exclusion problem. Parts I&II
- Lamport
- 1986
(Show Context)
Citation Context |