42,028 Hits in 4.4 sec

Synchronous consensus for dependent process failures

F.P. Junqueira, K. Marzullo
23rd International Conference on Distributed Computing Systems, 2003. Proceedings.  
To illustrate this abstraction, we consider Consensus for synchronous systems with both crash and arbitrary process failures.  ...  This abstraction models dependent process failures yet it is as simple to use as the Ø of Ò assumption.  ...  Voelker, André Barroso, Renata Teixeira, Dmitrii Zagorodnov, Xianan Zhang, and the people of RAMP (John Bellardo, Ranjita Bhagwan, Marvin McNett, Alper Mizrak, and Jessica Wang) for their support and comments  ... 
doi:10.1109/icdcs.2003.1203476 dblp:conf/icdcs/JunqueiraM03 fatcat:yo6pz3cqqfaixgfvnaohuhjsm4

Brief announcement

Fabrice Le Fessant, Philippe Raipin Parvédy, Michel Raynal
2003 Proceedings of the twenty-second annual symposium on Principles of distributed computing - PODC '03  
We have: • Crash failure model: both consensus and uniform consensus can be solved for any value of t (i.e., for t < n), • Omission failure model: consensus can be solved for any t, while uniform consensus  ...  So, from both theoretical and practical points of view, an attractive approach is to look for early-stopping consensus protocols, i.e., protocols whose number of rounds depends on f, the number of processes  ...  We have: • Crash failure model: both consensus and uniform consensus can be solved for any value of t (i.e., for t < n), • Omission failure model: consensus can be solved for any t, while uniform consensus  ... 
doi:10.1145/872035.872068 dblp:conf/podc/FessantPR03 fatcat:jwdrs5yitjajhb37m236ej6xpu

A Comprehensive Study on Failure Detectors of Distributed Systems

Bhavana Chaurasia, Anshul Verma
2020 Journal of scientific research  
Failure detectors are said to be unreliable because sometimes they suspect a correct process as a faulty process or they treat a faulty process as a correct process.  ...  The paper helps readers for the enhancement of knowledge about the basics of failure detectors and the different algorithms which are developed to solve the failure detection problems of distributed systems  ...  Enhancement Partially Crash failures, Detects the Abstract Uses polling concept, & al., 2004 algorithm for synchronous Consensus crashed properties also uses a logical ring unreliable failure systems processes  ... 
doi:10.37398/jsr.2020.640235 fatcat:znckxyrnnnf3npesjkjtifjde4

A New Algorithm to Solve Synchronous Consensus for Dependent Failures

Jun Wang, Min Song
2005 Sixth International Conference on Parallel and Distributed Computing Applications and Technologies (PDCAT'05)  
However, real systems may exhibit dependent failures. Cores and survivor sets are used to build an abstraction model for dependent process failures.  ...  In consensus problem, each process starts with a proposed initial value, and finally all non-faulty processes decide on the same value.  ...  Figure 1 : Synchronous consensus for dependent crash failures Our algorithm differentiates the processes of C from the rest of the processes in C − Π .  ... 
doi:10.1109/pdcat.2005.23 dblp:conf/pdcat/WangS05 fatcat:spndl5lvaffobjt5oblhowpm6q

Structured Derivation of Semi-Synchronous Algorithms [chapter]

Hagit Attiya, Fatemeh Borran, Martin Hutle, Zarko Milosevic, André Schiper
2011 Lecture Notes in Computer Science  
Separate implementations of TAB, withstanding different types of failures, allow to derive algorithms for consensus and set consensus under crash and omission failures.  ...  The paper presents a new communication primitive, Timely Announced Broadcast (TAB ), and uses it in algorithms for consensus and set consensus in the semi-synchronous model.  ...  In the synchronous model, exactly f /k + 1 rounds are required for solving k-set consensus in the presence of f crash failures, for any f < n [7, 8] .  ... 
doi:10.1007/978-3-642-24100-0_37 fatcat:bxz4hr22fjf3fbfrhnxeauu6tm

The consensus problem in unreliable distributed systems (a brief survey) [chapter]

Michael J. Fischer
1983 Lecture Notes in Computer Science  
Agreement problems involve a system of processes, some of which may be faulty. A fundamental problem of fault-tolerant distributed computing is for the reliable processes to reach a consensus.  ...  Acknowledgement The author is grateful for Ming Kao for help in assembling the bibliography and to Paul Hudak for many helpful comments on an early draft of this paper.  ...  The answer depends on whether synchronization upon termination is also required. Theorem 6 Let ¡ % $ $ 0 & .  ... 
doi:10.1007/3-540-12689-9_99 fatcat:mcno7yn4sjhjjbkib23ljjaqzy

The timed asynchronous distributed system model

F. Cristian, C. Fetzer
1999 IEEE Transactions on Parallel and Distributed Systems  
: synchronous system = a system that guarantees "certain communication" asynchronous system = a system which is not synchronous -In a synchronous system the frequency of failure is bounded (to achieve  ...  proposes a formal definition for the timed asynchronous distributed system model -presents measurements of process scheduling delays and hardware clocks drifts -Distributed systems can be classified as  ...  -how delta is chosen depends on the application's needs Datagram Failure Assumption: The asynchronous datagram service is assumed to have an omission / performance failure semantics.  ... 
doi:10.1109/71.774912 fatcat:7f6ak6ztbncwvcva2kotqqqkda

Perfect Failure Detection in the Partitioned Synchronous Distributed System Model

Raimundo José de Araújo Macêdo, Sérgio Gorender
2009 2009 International Conference on Availability, Reliability and Security  
In this paper we show that it is possible to implement a perfect failure detector P (one that detects all faulty processes if and only if those processes failed) in a non-synchronous distributed system  ...  Moreover, we show that even if strong partitioned synchrony is not valid, we are still able to take advantage of the existing synchronous partitions for improving the robustness of applications, by introducing  ...  RELATED WORK The solvability of FT problems, such as consensus [1] , depends decisively on the existence of guaranteed upper bounds for message transmission (δ) and process scheduling delays (φ) (i.e.  ... 
doi:10.1109/ares.2009.165 dblp:conf/IEEEares/MacedoG09 fatcat:k6tp5q77nfcczfdafss2mwgbq4

Synchrony and Time in Fault-Tolerant Distributed Algorithms [chapter]

Ulrich Schmid
2010 Lecture Notes in Computer Science  
process ID -There is a time after which every not yet crashed process outputs the same correct process p (the "leader") Implementability of FDs • If we can implement Ω [= weakest FD for consensus  ...  Shortcomings Θ-Model • We know that lock-step rounds are stronger than failure detectors that are sufficient for solving consensus: Replica consistency NEVER violated since indep. of δ !  ... 
doi:10.1007/978-3-642-15297-9_5 fatcat:as755mkxsffwvowk5xdz4mnhou

On the Impact of Fast Failure Detectors on Real-Time Fault-Tolerant Systems [chapter]

Marcos K. Aguilera, Gérard Le Lann, Sam Toueg
2002 Lecture Notes in Computer Science  
Specifically, we show how fast failure detectors can speed up consensus and fault-tolerant broadcasts, by providing fast algorithms and deriving some matching lower bounds, for synchronous systems with  ...  We investigate whether fast failure detectors can be useful -and if so by how much -in the design of real-time fault-tolerant systems.  ...  For instance, consensus requires in the worst-case time D(1 + f max ) in a synchronous system, where D is the maximum message delay and f max is the maximum number of process that may crash.  ... 
doi:10.1007/3-540-36108-1_24 fatcat:vz6r7u5jqjhqlh6erds7tcrnga

Failure Detection with Booting in Partially Synchronous Systems [chapter]

Josef Widder, Gérard Le Lann, Ulrich Schmid
2005 Lecture Notes in Computer Science  
In this paper, we present an alternative failure detector algorithm, which is based on a clock synchronization algorithm for the Θ-Model.  ...  Unreliable failure detectors are a well known means to enrich asynchronous distributed systems with time-free semantics that allow to solve consensus in the presence of crash failures.  ...  Moreover, the overhead can be reduced further by introducing local pauses between rounds (which does not even need timers as it can be done by counting suitable local events), see [2] for details.  ... 
doi:10.1007/11408901_3 fatcat:ln5w6eafpjgt7i26b2wxmxfy4q

Mastering agreement problems in distributed systems

M. Raynal, M. Singhal
2001 IEEE Software  
commitment Consensus Failure detection Point-to-point communication Figure 2.  ...  These authors focus on practical solutions for a well-known agreement problem-the nonblocking atomic commitment. 6 I E E E S O F T W A R E J u l y / A u g u s t 2 0 0 1 Applications Nonblocking atomic  ...  When the only possible failures are process crashes, the consensus and uniform consensus problems have relatively simple solutions in synchronous distributed systems.  ... 
doi:10.1109/ms.2001.936216 fatcat:tuptdvw7krhpbk6ckontljidsy

What Can be Computed in a Distributed System? [chapter]

Michel Raynal
2014 Lecture Notes in Computer Science  
The answer to this question depends on the environment in which evolves the considered distributed system, i.e., on the assumptions the system relies on.  ...  In the extreme case where the environment is such that there is no synchrony assumption and the computing entities may commit failures, many problems become impossible to solve (in these cases, a network  ...  He wants also to thank François Taïani for constructive comments.  ... 
doi:10.1007/978-3-642-54848-2_14 fatcat:clpsh3kdv5hh5cvlklobiodh3a

A Generic Group Communication Approach for Hybrid Distributed Systems [chapter]

Raimundo José de Araújo Macêdo, Allan Edgard Silva Freitas
2009 Lecture Notes in Computer Science  
Because of its integrated feature, our approach is capable of handling group communication for both synchronous and asynchronous distributed systems, dynamically adapting to the available QoS.  ...  Group Communication is a powerful abstraction that is being widely used to manage consistency problems in a variety of distributed system models, ranging from synchronous, to time-free asynchronous model  ...  For TimedCB, if failures are considered, the price is the same (the consensus price), no matter the number of tolerated failures (from 1 to n − 1).  ... 
doi:10.1007/978-3-642-02164-0_8 fatcat:ly5n2fu5lbbgvabe5vzp7dcmty

Harmful dogmas in fault tolerant distributed computing

Bernadette Charron-Bost, André Schiper
2007 ACM SIGACT News  
Consensus is a central problem in fault tolerant distributed computing. A vast number of (positive and negative) results for consensus in various system models have been established.  ...  thus yielding too restrictive solutions for real systems.  ...  This leads to the wrong message that synchrony (synchronous processes and synchronous links) is sufficient to achieve agreement: as shown in [17] , synchrony does not help for solving consensus in the  ... 
doi:10.1145/1233481.1233496 fatcat:z2movhmjzbccviwpyzwfdukx7i
« Previous Showing results 1 — 15 out of 42,028 results