Conference Paper

Bosco: One-Step Byzantine Asynchronous Consensus.

DOI: 10.1007/978-3-540-87779-0_30 Conference: Distributed Computing, 22nd International Symposium, DISC 2008, Arcachon, France, September 22-24, 2008. Proceedings
Source: DBLP

ABSTRACT Asynchronous Byzantine consensus algorithms are an important primitive for building Byzantine fault-tolerant systems. Algorithms for Byzantine consensus typically require at least two communication steps for decision; in many systems, this imposes a significant performance overhead. In this paper, we show that it is possible to design Byzantine fault-tolerant consensus algorithms that decide in one message latency under contention-free scenarios and still provide strong consistency guarantees when contention occurs. We define two variants of one-step asynchronous Byzantine consensus and show a lower bound on the number of processors needed for each. We present a Byzantine consensus algorithm, Bosco, for asynchronous networks that meets these bounds, even in the face of a strong network adversary.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Provide application processes with strong agreement guarantees despite failures is a fundamental problem of fault-tolerant distributed computing. Correct processes have not to be “polluted” by the erroneous behavior of faulty processes. This paper considers the consensus agreement problem in a setting where some processes can behave arbitrarily (Byzantine behavior). In such a context it is possible that Byzantine processes collude to direct the correct processes to decide on a “bad” value (a value proposed only by faulty processes). The paper has several contributions. It presents a family of consensus algorithms in which no bad value is ever decided by correct processes. These processes always decide a value they have proposed (and this is always the case when they all propose the same value) or a default value ?. These algorithms are called intrusion-free consensus algorithms. To that end, each consensus algorithm is based on an appropriate underlying broadcast algorithm. One of these abstractions, called validated broadcast is new and allows the design of a resilience-optimal consensus algorithm (i.e., it copes with up to t < n/3 faulty processes where n is the total number of processes). All proposed consensus algorithms assume the underlying system is enriched with additional computational power provided by a binary Byzantine consensus algorithm. The paper presents also a resilience-optimal randomized binary consensus algorithm based on the validated broadcast abstraction. An important feature of all these algorithms lies in the fact that they are signature-free (and hence particularly efficient).
  • [Show abstract] [Hide abstract]
    ABSTRACT: The paper considers the consensus problem in a partially synchronous system with Byzantine faults. It turns out that, in the partially synchronous system, all deterministic algorithms that solve consensus with Byzantine faults are leader-based. This is not the case of benign faults, which raises the following fundamental question: is it possible to design a deterministic Byzantine consensus algorithm for a partially synchronous system that is not leader-based? The paper gives a positive answer to this question, and presents a leader-free algorithm that is resilient-optimal and signature-free.
    Proceedings of the 11th international conference on Distributed computing and networking; 01/2010
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Among all classes of faults, Byzantine faults form the most general modeling of value faults. Traditionally, in the Byzantine fault model, faults are statically attributed to a set of up to t processes. This, however, implies that in this model a process at which a value fault occurs is forever “stigmatized” as being Byzantine, an assumption that might not be acceptable for long-lived systems, where processes need to be reintegrated after a fault.We thus consider a model where Byzantine processes can recover in a predefined recovery state, and show that consensus can be solved in such a model. Our model admits executions where over time every process is faulty as long as there are always enough correct processes.
    Theor. Comput. Sci. 01/2011; 412:4260-4272.

Full-text (2 Sources)

Available from
May 22, 2014