Open Access

ARTICLE

Aging Analysis Framework of Windows-Based Systems through Differential-Analysis of System Snapshots

Eun-Tae Jang¹, Sung Hoon Baek², Ki-Woong Park^1,*

1 SysCore Lab, Sejong University, Seoul, 05006, Korea
2 Department of Computer System Engineering, Jungwon University, Chungcheongbuk-do, 28024, Korea

* Corresponding Author: Ki-Woong Park. Email:

Computers, Materials & Continua 2022, 73(3), 5091-5102. https://doi.org/10.32604/cmc.2022.026663

Received 31 December 2021; Accepted 02 March 2022; Issue published 28 July 2022

Abstract

When a Windows-based system is used for an exceedingly long time, its performance degrades, and the error occurrence rate tends to increase. This is generally called system aging. To investigate the reasons for system aging, various studies have been conducted within the range of the operating system kernel to the user application. However, finding an accurate reason for system performance degradation remains challenging research topic. In this study, system monitoring was conducted by dividing a system into ‘before software installation,’ ‘after software installation,’ and ‘after software removal.’ We confirmed that when a software installed in a system is removed, various system elements, such as storage and memory, are not restored to the level prior to the software installation. Consequently, we established a hypothesis regarding the performance degradation of a computer system owing to repeated software installation/removal operations, investigated the correlation between system aging and repeated software installation/removal operations, and proposed a system aging analysis framework for analyzing the reason behind system aging. In the proposed system aging analysis framework, we aim to forcibly age a Windows-based system by repeating the software installation/removal operation by utilizing the system forced aging module. The framework identifies the elements affecting system performance through a differential data analysis of the system time-series data extracted by the system performance extraction and system component snapshot modules. Consequently, the aging analysis framework presented in this study is expected to be effectively utilized as an index for studying system aging.

---

Keywords

---