A Gaussian approximation of runtime estimation in a desktop grid project
Evsey V. Morozov, Олег Лукашенко, Alexander Rumyantsev, Evgeny Ivashko · 2017
We study a novel stochastic model describing the dynamics of a Desktop Grid project with many hosts and many workunits to be preformed. The model is based on the approximation of the basic process by means of the asymptotics of the superposed on-off sources both with light- and heavy-tailed distributions of the hosts' working sessions. Assuming that the amount of work to be done by the system is deterministic and finite, this approach leads to a Gaussian approximation of the process describing the summary processed work in the system. Using the properties of Gaussian processes and Monte-Carlo simulation, we estimate the expected runtime of the project.