Load Test Graphs/Reports

Load Test result statistics are made available in the form of graphs and reports. These help you in analyzing the performance of the web application under test.

You can view the dynamic graphs when play is in progress or view the detailed reports/graphs once the test execution is completed.

Following are some of the snapshots and description of the graphs and reports generated after the load test execution. To know the detailed description of the reports and graphs and how to analyze the same to identify potential bottlenecks, please refer to the white paper Performance Testing Report Analysis available in our web site in the following URL: http://www.adventnet.com/products/qengine/performance-testing-report-analysis.pdf

Graphs	Description	How to Read
Server Response Time vs Elapsed Time	This graph shows Elapsed time Vs time between Request sent to time of receiving last byte, in milliseconds.	-
Per URL Page Download Time	This graph shows the page response time on a per URL basis.	Bar graph showing the recorded pages on the X-axis and the response time (in milliseconds) on the Y-axis.
Page load time vs Elapsed Time	This graph shows Elapsed time Vs Page load time, in seconds.	Plots that stay close to the horizontal axis indicate better performance. This indicates the response time as seen by the user.
Transaction Status Summary. Request, Response Pending Details and Page Start and Complete Time	This graph shows the status of the number of pending requests to be sent and the pending responses to be received for the already sent requests. Also shows the Page Start Time and Complete Time.	Bar graph shows the transactions on the X-axis and Number of requests on the Y-axis. Appropriate tool tips are displayed upon pointing the mouse over the bar graph, to indicate the number of requests and responses pending. Hyperlinks are also provided to view the list of URLs for which the requests/responses are pending.
Active User Incrementation Vs Elapsed Time	This graph indicates the user addition interval over elapsed time.	Plots that stay close to the vertical axis to generate high loads in a short period.
Hits/sec vs Elapsed Time	This graph shows Elapsed time vs Number of requests handled successfully, per unit time.	In ramp-up mode , plots that stay close to the horizontal axis indicate better performance as load increases over time.
Error Percentage over Elapsed Time	This graph shows the cumulative number of errors recorded during the test on the vertical axis, with time (load) on the horizontal axis. These errors are typically connection refused or unexpected socket closure errors.	Plots that rise later in the test indicate better robustness.
Error Distribution Graph	This graph indicates the error status of each request placed	Bar graph on transaction basis
Active User Count, Error Percentage and Response Time over Elapsed Time to identify bottlenecks	This graph shows the Active User Count, Error Percentage and Response Time on the vertical axis and Elapsed Time on the horizontal axis. This graph enables you to identify the performance bottlenecks in your Web application.	Active User Count - Plots that stay close to the vertical axis to generate high loads in a short period. Error Percentage - Plots that rise later in the test indicate better robustness.
Throughput(bytes/sec) over Elapsed Time	This graph shows Elapsed time Vs Throughput in seconds	Plots that stay close to the vertical axis indicate better throughput.
Web Page Analyzer	This Report details about, Total Number of HTTP Request Total Number of Images Total Size of images Total Number of Javascript Total Size of Javascripts Total Number of Style sheet Total Size of Style Sheet Total Number of Socket Connections for Image/JS/CSS Total Size of the Web Page Apart from the above details, user can customize the report to add more properties to the report by writing their own implementation in java. Steps to write customization of web page analyzer report, 1. Write a Java class implementing com.adventnet.qengine.load.report.WebPageAnalyzerInterface 2. The above interface has generateReport method call taking path to the saved html source as parameter. 3. You can manipulate the HTML source and take the needed values and return it in Properties. 4. Properties will be obtained and reported in a table format below QEngine default Web Page Analyzer report. The Java class should be compiled under QEngine/jars folder having jars/QEngineLoadTest.jar in the classpath for compiling. And then restart the QEngine server to class file to load. Currently there is no UI to tell QEngine to generate Customized report. But this can be added to the conf\WebPageAnalyzer.conf file as said below, 1. Edit the conf\WebPageAnalyzer.conf 2.Add a new report node below "QEngine Default Report" and add "Class" as defined for the "QEngine Default Report". 3. Save and close the file.	This report provide report for each transaction separately. The value for each mentioned items will be shown in each line in the report. The customized properties also will be reported separately for each transaction.
Database Monitoring Graphs
Transaction Summary	This graph shows the total requests, total bytes sent and total bytes received.	Total requests, bytes sent and bytes received are plotted in the Y-axis and Elapsed time is plotted in the X-axis.
Connection Summary	This graph shows the number of open connections and the total number of aborted connections.	Total open connections and total aborted connections are plotted in the Y-axis and Elapsed time is plotted in the X-axis.
Server Monitoring Graphs
CPU % vs Time	This graph shows the cumulative CPU utilization over time	Plots that stay close to the horizontal axis indicate better performance.
Memory Usage vs Time	This graph shows the memory usage of the server over time	Plots that stay close to the horizontal axis indicate better performance.