An efficient delay-optimal distributed termination detection algorithm

  • Authors:

  • Nihar R. Mahapatra;Shantanu Dutt

  • Affiliations:

  • Department of Electrical & Computer Engineering, Michigan State University, East Lansing, MI 48824-1226, USA;Department of Electrical & Computer Engineering, University of Illinois at Chicago, 851 South Morgan Street (M/C 154), Chicago, IL 60607-7053, USA

  • Venue:
  • Journal of Parallel and Distributed Computing
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Distributed termination detection is a fundamental problem in parallel and distributed computing and numerous schemes with different performance characteristics have been proposed. These schemes, while being efficient with regard to one performance metric, prove to be inefficient in terms of other metrics. A significant drawback shared by all previous methods is that, on most popular topologies, they take Ω(P) time to detect and signal termination after its actual occurrence, where P is the total number of processing elements. Detection delay is arguably the most important metric to optimize, since it is directly related to the amount of idling of computing resources and to the delay in the utilization of results of the underlying computation. In this paper, we present a novel termination detection algorithm that is simultaneously optimal or near-optimal with respect to all relevant performance measures on any topology. In particular, our algorithm has a best-case detection delay of Θ(1) and a finite optimal worst-case detection delay on any topology equal in order terms to the time for an optimal one-to-all broadcast on that topology (which we accurately characterize for an arbitrary topology). On k-ary n-cube tori and meshes, the worst-case delay is Θ(D), where D is the diameter of the target topology. Further, our algorithm has message and computational complexities of Θ(MD+P) in the worst case and, for most applications, ΘM+P) in the average case-the same as other message-efficient algorithms, and an optimal space complexity of Θ(P), where M is the total number of messages used by the underlying computation. We also give a scheme using counters that greatly reduces the constant associated with the average message and computational complexities, but does not suffer from the counter-overflow problems of other schemes. Finally, unlike some previous schemes, our algorithm does not rely on first-in first-out (FIFO) ordering for message communication to work correctly.