On the minimal synchronism needed for distributed consensus
Journal of the ACM (JACM)
Concurrency control and recovery in database systems
Concurrency control and recovery in database systems
Adding time to synchronous process communications
IEEE Transactions on Computers - Special Issue on Real-Time Systems
Consensus in the presence of partial synchrony
Journal of the ACM (JACM)
Understanding fault-tolerant distributed systems
Communications of the ACM
IEEE Transactions on Computers
Impossibility of distributed consensus with one faulty process
Journal of the ACM (JACM)
Unreliable failure detectors for reliable distributed systems
Journal of the ACM (JACM)
Communications of the ACM
A nonblocking timed atomic commit protocol for distributed real-time database systems
Journal of Systems and Software
The weakest failure detector for solving consensus
Journal of the ACM (JACM)
Fail-awareness in timed asynchronous systems
PODC '96 Proceedings of the fifteenth annual ACM symposium on Principles of distributed computing
Non-blocking atomic commitment
Distributed systems (2nd Ed.)
Reaching Agreement in the Presence of Faults
Journal of the ACM (JACM)
SIGMOD '81 Proceedings of the 1981 ACM SIGMOD international conference on Management of data
Notes on Data Base Operating Systems
Operating Systems, An Advanced Course
Atomic Updates of Replicated Data
EDCC-2 Proceedings of the Second European Dependable Computing Conference on Dependable Computing
Another advantage of free choice (Extended Abstract): Completely asynchronous agreement protocols
PODC '83 Proceedings of the second annual ACM symposium on Principles of distributed computing
SFCS '83 Proceedings of the 24th Annual Symposium on Foundations of Computer Science
Conditions on input vectors for consensus solvability in asynchronous distributed systems
STOC '01 Proceedings of the thirty-third annual ACM symposium on Theory of computing
Conditions on input vectors for consensus solvability in asynchronous distributed systems
Journal of the ACM (JACM)
DISC'05 Proceedings of the 19th international conference on Distributed Computing
| Hi-index | 0.01 |
Allowing a set of objects to make a consistent decision, according to the values they propose, is a fundamental problem of distributed systems. When the system is reliable and synchronous the problem is simple. In the context of unreliable asynchronous systems that have to meet real-time constraints, the problem is far from trivial. A negative result stipulates that it is impossible to design a consensus protocol even if only one object crashes. The author considers timed asynchronous systems. He first proposes a general definition for decision-with-deadlines problems and then designs a protocol that solves these problems in timed asynchronous systems. An object can vote a or b. In a "good" configuration (there are not too many crashes and messages are timely) objects will decide consistently A or B according to the number of votes a or b that they have received. A and B are incompatible decision values. If there are too many crashes or too many messages that are not timely then it is possible that an object decides E (exception). Timed atomic commitment constitutes a particular instantiation of this problem.