Evaluating the impact of client network monitoring software on network path measurement

Pao Yue-kong Library Electronic Theses Database

Evaluating the impact of client network monitoring software on network path measurement

 

Author: Dou, Yuejiao
Title: Evaluating the impact of client network monitoring software on network path measurement
Degree: M.Sc.
Year: 2013
Subject: Computer networks -- Management.
Communication -- Network analysis.
Hong Kong Polytechnic University -- Dissertations
Department: Dept. of Computing
Pages: xiv, 124 leaves : ill. ; 30 cm.
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b2578653
URI: http://theses.lib.polyu.edu.hk/handle/200/6899
Abstract: OneProbe is a novel probing method for reliable and accurate metric-rich network path monitoring and measurement based on measurement metrics, such as round-trip time, forward path and reverse-path packet loss rates, and forward-path and reverse-path packet reordering. However, inaccurate measurement results may be introduced by running other software on the client machine. Therefore, it is important to evaluate the impact of client software on network path measurement. The overall objective of this dissertation is to investigate the root cause for the response time differences between the first response time and the second response time in OneProbe probing. After describing the considerations including testing methodology, testing software tools and testing schedule on a testbed, the author sets up a testbed including a client, a server as OneProbe measurement server and a DAG (Data Acquisition and Generation) device guarantees 100% packet capture, to investigate the causes of the response time differences on client by a combination of round-trip probes (RTPs) and two-way probes (TWPs) probing in OneProbe. The experiments on the testbed are repeated on client by using different variables (one at a time), such as different brands of security software, operating systems, or network cards. Then, the packet timing diagrams on the client are compared among all the combinations with different variables, in order to seek for a conclusive result to ascertain the differences of impacts on the first and second response time, and to reveal the main causes of the discrepancy of them, especially for the first one is greater than the second one in most situations.
The most important contents of this dissertation are presented in Chapter 4 to Chapter 6. The related test suites on the difference of response time on client in RTP and TWP are divided into three test suites for three different testing situations on Windows, Linux and network interface card respectively. The first test suite focuses on testing for the effects of Windows Client for different brands of security software in serial test cases. The second one emphasizes on testing and analysis for effect of Linux Client for different brands of security software according to serial test cases, and comparisons between the test results from Linux Client and Windows Client have been illustrated in detail. Finally, the third one is based on testing for the effect of different network interface cards on the same client. As author's contributions, some important conclusions are gotten as following: Firstly, on Windows and Linux Client: different brands of security software, browsers, operating systems and network cards have impacts on the values of the first and second response time, and there are some discrepancies between them. Secondly, for different probing methods on Windows Client and Linux Client: in RTP and TWP, results show that the first response time is greater than the second one in most probe pairs. The delays in TWP are greater than that in RTP because the client node response until the second probe packet arrives. Eventually, as a useful suggestion, for future network path measurement, we should take care of the different impacts on response time in RTP and TWP. Moreover, the measurement impacts from the security software might influence test results while using OneProbe, which may be used to imply what kinds of security software running on clients for future study. As time is limited, only several security software, operating systems and network cards have been tested. More tests should be done on more variables in future.

Files in this item

Files Size Format
b25786532.pdf 2.290Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

     

Quick Search

Browse

More Information