ZebraTester Diagrams/Graphs/Charts

Here’s a brief description of a number of diagrams that are available, just based on one selected metric, Page Response Time.

Diagram: Page Response Time

Detail Per Page

The Detail Per Page option shows the Response Time per Page detail for each load-tested page.

Average/90th Percentile (Per Page)

Average/90% shows a bar for each page for each. This diagram also shows the Avg/90% response time progression of the web pages.

Results per URL Call (Overview)

Displays statistics about all URL calls.

Columns

Test: consecutively numbered. Clicking on a number displays the URL detail menu

# Passed: total number of successful calls

# Failed: total number of failed calls. If this value is greater than zero, you can click on it to display the corresponding error snapshots

AV Time: average response time per URL call, or per web page

⇐ 90 %: slowest response time within the fastest 90% of all measured values (90% percentile value). This result is only available if the response time has been collected at least 5 times, depending on the percentile sampling rate which was selected when the test run was started. Clicking on this value displays the corresponding response time percentile diagram

AV Size: the average size of transmitted and received data per URL call, or per web page

URL: the URL called

Response Content Throughput / In-Depth Measurement of HTTP(S) Response-Streams

The in-depth measurement of HTTP(S) response-streams is only available if you have to enable the additional option "Resp Throughput Chart per Call" as part of the "Additional sampling rate by URLcall" when starting the load test. Furthermore, you should configure a "maximum acceptable response time" in order that ZebraTester can calculate and compare the necessary network throughput.

This feature is especially useful for Web pages that contain videos and allows us to detect if jerky video playback occurs during the viewing of a video, respectively to diagnose if enough network bandwidth is available for all simulated users so that the video can be viewed by each user without interruption. However, this feature can also commonly used as a reference for the optimization of any response data. The corresponding charts are showing in different colors the times elapsed for receiving fragments of user data (in red color) and the times elapsed for receiving the overhead data of the chunked protocol (in blue color).

Measured internal throughput of a video on a preset viewing time of 3 minutes (180,000 milliseconds).

The linear flow and the flow rate peak at the beginning of receiving the data indicate that the delivery is made by a special video server which prevents on the one hand network peaks and ensures on the other hand that no jerky video playback occurs.

Throughput measurement of a PDF document which should be received in 30 seconds by a linear network throughput, in order that the beginning of the document can already be viewed after a few seconds. The second measured sample does not meet this requirement.

Throughput measurement of an HTML response received from a web portal server. It is conspicuous that the most response time is spent in the chunked protocol overhead, but that the user data (payload) is received in a relatively short time.

One explanation could be that the Web page is "calculated" piece by piece by the portal server (page navigation, page main content, page footer), and that some server-internal delay times occurred during the calculations.

Results per URL Call (Details)

Displays measurement details about URL calls. If this menu is invoked from the URL overview menu by clicking on a test number, the selected URL call is marked with a blue background.

Columns

Test: consecutively numbered. Clicking on a number displays the URL overview menu

Av Net Con: average time per URL call required to open a network connection to the webserver, before HTTP data are sent (socket open time). If the HTTP protocol option keep alive is supported by the webserver, this time is only measured for some URL calls - and not on all - because the network connections have been reused

Av Req Trm: average time per URL call to transmit the HTTP request data to the web server, measured after the network connection has been opened to the webserver

Av Wait: average time, per URL call, waiting for the first byte of the HTTP response (-header) from the webserver, measured after the HTTP request data have been transmitted

Av Header Rcv: average time, per URL call, receiving the HTTP response header from the web server, measured after the first byte has been received

Av Content Rcv: average time, per URL call, receiving the HTTP response content from the webserver (HTML data, images, etc.), measured after the HTTP response header has been received

Min Time: smallest-ever measured time of the URL call (request + response)

Av Time: average time of the URL call (request + response)

Max Time: highest-ever measured time of the URL call (request + response)

AV Size: the average size of transmitted and received data per URL call, or per web page

Av Throughput: average network throughput per URL call (request + response)

Diagram: Response Time Percentiles

This screen contains, per web page and per URL, the response time percentile diagram. These diagrams display a cumulative statistical distribution of response times but are only available if an Additional Sampling Rate per Page Call and/or an Additional Sampling Rate per URL Call option was set when starting the test run, and at least 5 individual measurements have been collected during the test run.

By using the option lists, you can select the web page and - within the page - the URL. The option “—“ for a URL means that the percentile diagram for the web page is displayed (instead of a specific URL of the web page).

“Cumulative statistical distribution” means that only the slowest URL call, within a percentage of all fastest URL calls, is flowing inside the curve. For example, 95% means that 95% of all URL calls have a response time faster than, or equal to, the shown value.

The collected individual measurements can be displayed by clicking on the Apply button. It is also possible to export the individual measurements in the form of an HTML table.

Diagram: Top Time-Consuming URLs

Shows a compilation of the slowest URLs (average and 90% percentile response time values) and the response time distribution of the slowest URL per page (for each page), and the response time per media-type (text/html, image/gif …).

Diagram: Concurrent Users

Shows the number of users during the test run. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram: Session Time

Shows the response time per successfully-executed loop (repetition of a web surfing session) during the test run. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

The accumulated user’s think time of the loop is shown by a red line.

Diagram: Web Transaction Rate

Shows the number of successfully-executed URL calls per second (hits per second) during the test run, measured over all simulated users. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram Users Waiting for Response

Shows the number of users who are waiting for a response from the webserver, measured over all simulated users. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram: Completed Loops

Shows the number of successfully-completed web surfing sessions (loops) per minute

measured over all concurrent users. The number of data points depends on the

Statistic Sampling Interval which was set when the test run was started.

Diagram: TCP Socket Connect Time

Shows the time to open a new network connection to the webserver before data are sent (socket open time). The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram: Network Throughput

Shows the total network throughput of the test run, measured over all users. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram: HTTP Keep-Alive Efficiency

Shows the efficiency of the HTTP keep-alive protocol option (percentage of reused network connections), measured over all users and URL calls.

Diagram: SSL Cache Efficiency

Shows the efficiency of the client-side SSL session cache, which depends on the web server SSL configuration. In other words, this shows the percentage of abbreviated SSL handshakes, measured over all users. This diagram is only available when each user has executed at least 5 successful loops, and when the encrypted HTTPS protocol has been used.

Diagram: Session Failures

Shows the number of failed web surfing session (failed loops) which occurred during the test run. The number of data points depends on the Statistic Sampling Interval which was set when the test run was started.

Diagram: Error Types

Shows a compilation of the most frequently-occurring error types. Note: this basic error information is accurately measured, also in the case when not enough memory was left to capture error snapshots for all occurred errors.

Diagram: Number of Errors per Page

Shows a compilation of the web pages which experienced the most errors.

Diagram: Number of Errors per URL

Shows a compilation of the URLs which experienced the most errors.