International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 07 | July 2023
p-ISSN: 2395-0072
www.irjet.net
Benchmarking Techniques for Performance Analysis of Operating Systems and Programs Devansh Baid1, Tanmay Mandal2, Nilophar Joglekar3, Akila Victor4 123Student, School of Computer Science and Engineering, VIT University, Tamil Nadu, India
4Associate Professor, School of Computer Science and Engineering, VIT University, Tamil Nadu, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Benchmarking computer systems is an
is only interested in scientific computing. For scientific computing, a wide range of high-performance computers are available, ranging from vector computers with a small number of processors sharing common memory to machines with thousands of processors and distributed memories. The performance range of these computers can vary significantly, perhaps by a factor of a thousand or more, depending on how well the problem and software fit the underlying architecture and operating system.
important, albeit time-consuming, process that provides insight into a system's performance, exposes weaknesses, and permits comparison between systems or versions. Current UNIX operating system benchmarking programs are heavily weighted toward CPU and hardware performance. As operating systems become more complicated, so do the programs that run on them. The commonly accessible benchmarks currently do not take this into account, making them kindly unrealistic and uninformative. We give a comparison of extensive and widely used operating systems. Eventually, a detailed and comprehensive conclusion has been reached based on the results of various tests.
2. RELATED WORK There are numerous existing technologies for performance evaluation and benchmarking tools. Computer system designers, administrators, and users are interested in performance evaluation since their goal is to achieve or deliver the best performance at the lowest possible cost. As system performance varies greatly from one application area to the next, no one statistic can be used to assess computer system performance across all applications. Loads on various system components also have a significant impact on performance.
Key Words: Benchmarking; Operating Systems, Process Management, Task Scheduling, File Management, Computer Performance, Network Monitor
1. INTRODUCTION In the early days of computing, the primary purpose of a programmer was to create a workable program with little regard for its efficiency. Von Neumann contrasted the speed with which early computers (including the ENIAC) conducted multiplication when computing ballistic trajectories in 1946 [McK88]. Herbst et al. measured the instruction mix of Maniac computer programs in 1955.
The performance of a computer system for a specific task can be measured using a variety of performance assessment and benchmarking tools. However, very few benchmarks, if any, examine how a system performs under various system loads. Most of the benchmarks are intended to run on an 'idle system'. As a result, they provide a measurement of a system's "peak efficiency" under a specific category of workload.
Performance evaluation is of importance to computer system designers, administrators, and users alike. Designers analyze several alternative designs to select the finest one. Administrators compare various systems to determine the optimal system for a collection of applications. Users examine many installed systems to determine which system is best suited for a certain task. The key objective in computer system design, procurement, and use is to provide maximum performance at the lowest possible cost.
In the world of computing or computer networking, the benchmarking idea is not new. When one talks about "benchmarking tools," they typically mean a program or set of programs that are used to compare one solution's performance to that of another under specific reference circumstances. Benchmarking methods have been employed to rate the efficiency of computers and computer networks since the 1970s.
The evaluation of a computer system's performance is a challenging task. When people hear the word "performance," they may conjure up entirely different images. persons who work with huge databases tend to think of performance in terms of transactions per second, but persons in scientific and technological fields may be interested in the number of floating-point operations per second. Even with a small emphasis, evaluating performance is not simple. Assume one
© 2023, IRJET
|
Impact Factor value: 8.226
In the research projects CREW and OneLab2, Benchmarking of wired and wireless computer networks is presented as a study topic and the BonFIRE experiment investigates benchmarking of applications and virtual machines in an Infrastructure-as-a-Service (IaaS) setting.
|
ISO 9001:2008 Certified Journal
|
Page 1205