Informative parameters of the IT-infrastructure: Determination experience

Abstract

The paper is devoted to the problem of informative parameters choice to monitor and optimize IT infrastructure (ITI) in organizations. The purpose of the research is to reduce number of parameters for monitoring a complex system on the example of ITI of the corporation Tyumenneftegeophysica. To censor the data and estimate normality of experiment matrix columns we observed ITI in stationary conditions, measured parameters and formed the experiment matrix 1344×4397. Spearman's correlation analysis was used. As a result, 581 mutually independent indicators were defined as informative out of 4397 measurable ones. If necessary, the values of other parameters can be estimated from the values of informative ones. Using the suggested methodology will allow researchers and designers to obtain a subset of informative parameters quickly to monitor an object.