### Citations

850 | Wait-Free Synchronization
- Herlihy
- 1991
(Show Context)
Citation Context ... analyse here the results he gave regarding two particular types of sharedmemory systems, namely the ones in which processes communicate via SingleWriter/MultiReader atomic registers (SWMR for short) =-=[10, 12, 11]-=-, and Atomic-Snaphot objects [1, 2], respectively. For the first type, each process p is associated with a read/write register Rp such that (i) p is the sole process that is allowed to write into Rp, ... |

651 | Reaching Agreement in the Presence of Faults
- Pease, Shostak, et al.
- 1980
(Show Context)
Citation Context ...how it is possible to compare and hierarchize such predicates. In [4], Charron-Bost et al. generalized the HO model to cope with value failures. This extended model covers both the Byzantine failures =-=[13]-=- and the dynamic transmission faults of [14]. The HO model thus appears to be suitable for systems with any type of failures. Future works may try to apply the techniques presented in this paper to ri... |

183 | Atomic snapshots of shared memory
- Afek, Attiya, et al.
- 1993
(Show Context)
Citation Context ...ng two particular types of sharedmemory systems, namely the ones in which processes communicate via SingleWriter/MultiReader atomic registers (SWMR for short) [10, 12, 11], and Atomic-Snaphot objects =-=[1, 2]-=-, respectively. For the first type, each process p is associated with a read/write register Rp such that (i) p is the sole process that is allowed to write into Rp, and (ii) every process can read the... |

93 | Efficient Synchronization on Multiprocessors with Shared Memory
- Kruskal, Rudolph, et al.
- 1988
(Show Context)
Citation Context ... analyse here the results he gave regarding two particular types of sharedmemory systems, namely the ones in which processes communicate via SingleWriter/MultiReader atomic registers (SWMR for short) =-=[10, 12, 11]-=-, and Atomic-Snaphot objects [1, 2], respectively. For the first type, each process p is associated with a read/write register Rp such that (i) p is the sole process that is allowed to write into Rp, ... |

65 |
On Inter-process Communication. Part II: Algorithms
- Lamport
- 1986
(Show Context)
Citation Context ... analyse here the results he gave regarding two particular types of sharedmemory systems, namely the ones in which processes communicate via SingleWriter/MultiReader atomic registers (SWMR for short) =-=[10, 12, 11]-=-, and Atomic-Snaphot objects [1, 2], respectively. For the first type, each process p is associated with a read/write register Rp such that (i) p is the sole process that is allowed to write into Rp, ... |

61 |
Time is not a healer
- Santoro, Widmayer
- 1989
(Show Context)
Citation Context ...ree and failure model are encapsulated into the same abstract entity. In [6], Charron-Bost and Schiper combined this approach with the Transmission Faults model, introduced by Santoro and Widmayer in =-=[14]-=-, designed for synchronous message-passing systems and which locates the failure without specifying their cause. As a result, Charron-Bost and Schiper developed the Heard-Of model (HO for short), a co... |

53 | Round-by-round fault detectors: unifying synchrony and asynchrony
- Gafni
- 1998
(Show Context)
Citation Context ...ound-based model, fault-tolerant distributed systems, shared-memory model, Single-Writer/Multi-Reader atomic registers, Atomic-Snapshot registers.Introduction hal-00294595, version 1 - 9 Jul 2008 In =-=[9]-=-, Gafni defined a new round-based computational model, called Round-by-Round Fault Detector model (RRFD for short), for the analysis of distributed systems. In this model, computations are defined to ... |

47 |
The Heard-Of model: computing in distributed systems with benign faults,
- Charron-Bost, Schiper
- 2009
(Show Context)
Citation Context ...each process p, the module provides a set of suspected processes from which p will not wait for a message. Hence, synchrony degree and failure model are encapsulated into the same abstract entity. In =-=[6]-=-, Charron-Bost and Schiper combined this approach with the Transmission Faults model, introduced by Santoro and Widmayer in [14], designed for synchronous message-passing systems and which locates the... |

37 | A Simple Algorithmically Reasoned Characterization of Wait-free Computations
- Borowsky
- 1997
(Show Context)
Citation Context ...one. Hence, at the end of round r, we have either HOp ⊆ HOq or HOq ⊆ HOp. ✷ The fact that an Atomic-Snaphot object can be implemented in a system whose executions satisfy PRD is a simple corollary of =-=[5]-=-. Combining this result with Proposition 2.1, we derive the following theorem: p∈Π Theorem 2.2. The communication predicate PRD entirely captures the communication properties of Atomic-Snapshot system... |

32 |
Distributed Computing
- Attiya, Welch
- 2004
(Show Context)
Citation Context ...t change for a given number of rounds, then p defines NewHOp to be the set of all processes it heard of during ρ. This scheme is inspired by the double-collect used in many implementations (e.g. [2], =-=[3]-=-) of Atomic-Snapshot objects by SWMR atomic registers. 6Algorithm 3 The SRD algorithm 1: Initialization: 2: Dp ⊆ Π; initially {p} 3: NewHOp ⊆ Π; initially ∅ 4: Knownp ⊆ Π; initially ∅ 5: k, Np ∈ N; i... |

10 | How hard is it to take a snapshot?
- Fich
- 2005
(Show Context)
Citation Context ...ponding to Atomic-Snapshot systems in Section 2, and demonstrate that the one he proposed for SWMR systems is strictly weaker. The well-known equivalence between SWMR systems and Atomic-Snaphots ones =-=[1, 2, 8]-=- then leads us to consider an alternative communication predicate. In Section 3, we prove this third predicate to be (1) implementable in SWMR systems and (2) equivalent to the one corresponding to At... |

1 |
André Schiper. Tolerating corrupted communication
- Biely, Widder, et al.
- 2007
(Show Context)
Citation Context ...aracterization of classical models only in terms of predicates that capture the properties of their communications. Moreover, we show how it is possible to compare and hierarchize such predicates. In =-=[4]-=-, Charron-Bost et al. generalized the HO model to cope with value failures. This extended model covers both the Byzantine failures [13] and the dynamic transmission faults of [14]. The HO model thus a... |