This article is cited in 1 paper

A Gaussian approximation of the distributed computing process

O. V. Lukashenko^ab, E. V. Morozov^ab, M. Pagano<sup>c

a</sup> Institute of Applied Mathematical Research of Karelian Research Centre of RAS, 11 Pushkinskaya Str., Petrozavodsk 185910, Republic of Karelia, Russian Federation
^b Petrozavodsk State University, 33 Lenin Str., Petrozavodsk 185910, Republic of Karelia, Russian Federation
^c University of Pisa, 43 Lungarno Pacinotti, Pisa 56126, Italy

Abstract: The authors propose a refinement of the stochastic model describing the dynamics of the Desktop Grid (DG) project with many hosts and many workunits to be performed, originally proposed by Morozov et al. in 2017. The target performance measure is the mean duration of the runtime of the project. To this end, the authors derive an asymptotic expression for the amount of the accumulated work to be done by means of limit theorems for superposed on-off sources that lead to a Gaussian approximation. In more detail, depending on the distribution of active and idle periods, Brownian or fractional Brownian processes are obtained. The authors present the analytic results related to the hitting time of the considered processes (including the case in which the overall amount of work is only known in a probabilistic way), and highlight how the runtime tail distribution could be estimated by simulation. Taking advantage of the properties of Gaussian processes and the Conditional Monte-Carlo (CMC) approach, the authors present a theoretical framework for evaluating the runtime tail distribution.

Keywords: Gaussian approximation, distributed computing, fractional Brownian motion.

Received: 15.04.2019

Language: English

DOI: 10.14357/19922264190215

Bibliographic databases:

•