Determination of specialized software API dependencies under maximum possible load without significant system degradation

Objective. The article considers the topic of determining the dependencies of the API of specialized software under the maximum possible load without significant system degradation.

Method. The study is based on regression analysis methods. As part of the research, limitations on the amount of random access memory and the number of processor cores are considered, which allows us to study the behavior and identify bottlenecks of SSS in virtualized environments with limited resources.

Result. To analyze the dependencies of expected pairs of parameters, it is proposed to use the following models: linear, exponential, logarithmic and polynomial (up to 4th degree). The best model is selected based on the adjusted coefficient of determination, with optional outlier filtering using the interquartile range method and a test of the statistical significance of the regression equation. Additionally, sample variance is calculated to assess the quality of the model fit.

Conclusion. This methodology presents a systematic approach to analyzing API performance dependencies in specialized software, based on the principles of applied statistics and engineering interpretation. It can be used to make informed decisions about reducing the scope of load testing. The methodology enables the translation of statistical patterns into engineering actions: infrastructure optimization, code refactoring, and scaling.

1. Chernyaeva S.N., L.A. Korobova, Yu.V. Bugaev. Algorithm of collective choice on a set of alternatives that is boundless for the decision maker. Engineering technologies. 2024;1(5): 22-27. - EDN JMQXRJ.

2. Gmurman, V.E. Probability theory and mathematical statistics: textbook. - 12th ed., revised and enlarged. - Moscow: Higher education, 2006; 479 p.

3. Chernyaeva S.N., Korobova L.A. Analysis of the degeneracy of expert ordering matrices in collective choice problems. Modeling of energy-information processes: Collection of articles of the XII national scientific and practical conference with international participation. - Voronezh: Voronezh State University of Engineering Technologies, 2024; 213-224. - EDN IDJAIF.

4. Kremer N.Sh., Putko B.A. Econometrics. - Moscow: UNITY-DANA, 2002. - 315 p.

5. Ivliev, M.N. Financial Mathematics. Methods and Models in Economics. Collection of problems: textbook / M.N. Ivliev, L.A. Korobova, K.V. Chekudaev. – Voronezh: Voronezh State University of Engineering Technologies, 2019; 91 p. – ISBN 978-5-00032-444-8. – EDN MHQVNZ.

6. Korchuganova M.A. Laboratory practical training in the discipline "Econometrics": Tomsk: Publishing house of Tomsk Polytechnic University, 2016; 91 p.

7. Implementation of the Extrapolation Method of Expert Assessments in Selection Problems / S. Chernyaeva, L. Korobova, M. Ivliev [et al.] High-Performance Computing Systems and Technologies in Scientific Research, Automation of Control and Production, Barnaul, May 15-16, 2020 / Vladimir Jordan Nikolay Filimonov Ilya Tarasov Vladimir Faerman (Eds.) – Barnaul: Springer, 2020;1304: 47-161. – DOI 10.1007/978- 3-030-66895-2_10. – EDN HINSXG.

8. Khnykin, D.E. Methodology and results of applying reverse engineering / D.E. Khnykin, L.A. Korobova Engineering technologies. 2025;1(9):30-35. – EDN UVTHPV.

9. Bugayev Yu.V., Korobova L.A., Shurupova I.Yu. On the statistical stability of the optimal solution found by the regression equation Bulletin of the Voronezh State University of Engineering Technologies. 2024;86(2(100):48-55. – DOI 10.20914/2310-1202-2024-2-48-55. – EDN IYNADI.