Operating system efficiency evaluation on the base of measurements analysis with the use of non-extensive statistics elements

Paweł Dymora,Kamil Zelazny,Mirosław Mazurek

doi:10.2478/umcsinfo-2014-0015

Abstract

AbstractThe major goal of this article was to evaluate the efficiency of Linux operating system using statistical self-similarity and multifractal analysis. In order to collect the necessary data, the tools available in Linux such as vmstat, top and iostat were used. The measurement data collected with those tools had to be converted into a format acceptable by applications which analyze statistical selfsimilarity and multifractal spectra. Measurements collected while using the MySQL database system in a host operating system were therefore analyzed with the use of statistical self-similarity and allowed to determine the occurrence of long-range dependencies. Those dependencies were analyzed with the use of adequately graduated diagrams. Multifractal analysis was conducted with the help of FracLab application. Two methods were applied to determine the multifractal spectra. The obtained spectra were analyzed in order to establish the multifractal dependencies.

Highlights

Computer systems can be divided into two categories
The first category is defined as simple systems and the second – complex systems
As far as the definition of the complex systems is con-cerned many different explanations of the term are found in technical literature [1, 2]

Summary

Introduction

Computer systems can be divided into two categories. The first category is defined as simple systems and the second – complex systems. As far as the definition of the complex systems is con-cerned many different explanations of the term are found in technical literature [1, 2]. It is important to notice that complex. Systems cannot be described and characterized just by one simple definition. The main characteristics of complex systems are: common application structures and statistical self-similarity in time and space [3, 4]. The efficiency planning is dynamic, controlling processes and managing systems resources are closely related and there is no guarantee of the service quality

Statistical self-similarity

Multifractal spectra

Operating system efficiency measurements and statistical analysis

Multifractal analysis

Conclusions