Multistage queueing model of the distributed data processing system

Authors: Grachev V. V., Moiseev A. N., Nazarov A. A., Yampolskiy V. Z.

Annotation: Mathematical model for the technical system of distributed data processing is considered in the paper. The model is presented in a form of the multistage queueing system. High intensity of incoming queries is a typical feature of the considering system. It was shown that probability distribution of event arrived during fixed time period can be approximated by normal probability distribution for the most of studied homogeneous event processes. Multistage queueing system with unlimited servers count in each stage, general time of serving and high intensive input flow is considered in the paper also. It was shown that multidimensional probability distribution of busy servers count at each stage of the system can be approximated by multidimensional normal probability distribution. Parameters of this distribution were obtained in the paper.

Keywords: distributed data processing system, multistage queueing system, high-intensive arrival process