The weakest failure detectors to solve certain fundamental problems in distributed computing

Delporte-Gallet, Carole; Fauconnier, Hugues; Guerraoui, Rachid; Hadzilacos, Vassos; Kouznetsov, Petr; Toueg, Sam

doi:10.1145/1011767.1011818

Cited by 89 publications

(108 citation statements)

References 26 publications

(64 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Recall that ♦S is the weakest to solve the problem only in majority-correct environments [29]. In environments where an arbitrary number of processes may crash, the weakest failure detector for the problem is a stronger oracle (♦S, Σ) [15]. Given that ♦S encapsulates some fairness constraints, and Σ can be implemented in an asynchronous system with majority correct, we conjecture that Σ and majority-correct encapsulate equivalent fairness constraints in the system.…”

Section: Open Questionsmentioning

confidence: 99%

“…For example, consider wait-free consensus. The weakest failure detector for this problem in asynchronous shared-memory systems is Ω [29] whereas the weakest failure detector to solve the same problem in message-passing systems is (Ω, Σ); 1 that is, process have access to two failure detectors Ω and Σ [15]. Similarly, for solving wait-free k-set agreement, the weakest failure detector in shared-memory systems is anti-Ω k 2 [23], but the weakest failure detector in message-passing systems remains an open problem [8,9].…”

Section: Weakest Models For Failure Detectors In Shared Memorymentioning

confidence: 99%

“…In fact, an asynchronous shared-memory system is equivalent to an asynchronous system under message passing that is augmented with the quorum failure detector Σ [15]. Consequently, the results from [38] need not (and do not) carry over to message-passing systems.…”

Section: Weakest Models For Failure Detectors In Shared Memorymentioning

confidence: 99%

See 2 more Smart Citations

Failure Detectors Encapsulate Fairness

Pike

Sastry

Welch

2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Failure detectors have long been viewed as abstractions for the synchronism present in distributed system models. However, investigations into the exact amount of synchronism encapsulated by a given failure detector have met with limited success. The reason for this is that traditionally, models of partial synchrony are specified with respect to real time, but failure detectors do not encapsulate real time. Instead, we argue that failure detectors encapsulate the fairness in computation and communication. Fairness is a measure of the number of steps executed by one process relative either to the number of steps taken by another process or relative to the duration for which a message is in transit. We argue that failure detectors are substitutable for the fairness properties (rather than real-time properties) of partially synchronous systems. We propose four fairness-based models of partial synchrony and demonstrate that they are, in fact, the 'weakest system models' to implement the canonical failure detectors from the Chandra-Toueg hierarchy. We also propose a set of fairness-based models which encapsulate the G c parametric failure detectors which eventually and permanently suspect crashed processes, and eventually and permanently trust some fixed set of c correct processes.

show abstract

Section: Open Questionsmentioning

confidence: 99%

Section: Weakest Models For Failure Detectors In Shared Memorymentioning

confidence: 99%

See 1 more Smart Citation

Failure Detectors Encapsulate Fairness

Pike

Sastry

Welch

2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…It has been shown that the failure detector class denoted Ω is the weakest failure detector class that allows consensus to be solved in message-passing asynchronous systems where a majority of processes never crash [9]. It has also been shown that the pair (Σ, Ω) is the weakest failure detector class when any number of processes may crash [14,15]. (These failure detector classes are precisely defined later in the paper.…”

Section: Introductionmentioning

confidence: 99%

Anonymous asynchronous systems: the case of failure detectors

Bonnet

Raynal

2012

Distrib. Comput.

View full text Add to dashboard Cite

Due the multiplicity of loci of control, a main issue distributed systems have to cope with lies in the uncertainty on the system state created by the adversaries that are asynchrony, failures, dynamicity, mobility, etc. Considering message-passing systems, this paper considers the uncertainty created by the net effect of three of these adversaries, namely, asynchrony, failures, and anonymity. This means that, in addition to be asynchronous and crash-prone, the processes have no identity.Trivially, agreement problems (e.g., consensus) that cannot be solved in presence of asynchrony and failures cannot be solved either when adding anonymity. The paper consequently proposes anonymous failure detectors to circumvent these impossibilities. It has several contributions. First it presents three classes of failure detectors (denoted AP , AΩ and AΣ) and show that they are the anonymous counterparts of the classes of perfect failure detectors, eventual leader failure detectors and quorum failure detectors, respectively. The class AΣ is new and showing it is the anonymous counterpart of the class Σ is not trivial. Then, the paper presents and proves correct a genuinely anonymous consensus algorithm based on the pair of anonymous failure detector classes (AΩ, AΣ) ("genuinely" means that, not only processes have no identity, but no process is aware of the total number of processes). This new algorithm is not a "straightforward extension" of an algorithm designed for non-anonymous systems. To benefit from AΣ, it uses a novel message exchange pattern where each phase of every round is made up of sub-rounds in which appropriate control information is exchanged. Finally, the paper discusses the notions of failure detector class hierarchy and weakest failure detector class for a given problem in the context of anonymous systems.

show abstract

“…Hence, the information provided about the failure of a process might not be perfect. It is common to ask what information about failures is necessary and sufficient to circumvent some specific impossibility, e.g., consensus [3], atomic commit [6], mutual exclusion [7], etc.…”

Section: Introductionmentioning

confidence: 99%

On the weakest failure detector ever

et al. 2009

Self Cite

View full text Add to dashboard Cite

Many problems in distributed computing are impossible when no information about process failures is available. It is common to ask what information about failures is necessary and sufficient to circumvent some specific impossibility, e.g., consensus, atomic commit, mutual exclusion, etc. This paper asks what information about failures is needed to circumvent any impossibility and sufficient to circumvent some impossibility. In other words, what is the minimal yet non-trivial failure information.We present an abstraction, denoted Υ, that provides very little failure information. In every run of the distributed system, Υ eventually informs the processes that some set of processes in the system cannot be the set of correct processes in that run. Although seemingly weak, for it might provide random information for an arbitrarily long period of time, and it only excludes one possibility of correct set among many, Υ still captures non-trivial failure information. We show that Υ is sufficient to circumvent the fundamental wait-free set-agreement impossibility. While doing so, we (a) disprove previous conjectures about the weakest failure detector to solve set-agreement and we (b) prove that solving set-agreement with registers is strictly weaker than solving n + 1-process consensus using n-process consensus.We prove that Υ is, in a precise sense, minimal to circumvent any wait-free impossibility. Roughly, we show that Υ is the weakest eventually stable failure detector to circumvent

show abstract

The weakest failure detectors to solve certain fundamental problems in distributed computing

Cited by 89 publications

References 26 publications

Failure Detectors Encapsulate Fairness

Failure Detectors Encapsulate Fairness

Anonymous asynchronous systems: the case of failure detectors

On the weakest failure detector ever

Contact Info

Product

Resources

About