Running Load Tests

Once you’ve created a session of Http Requests you can then start a load test run by specifying:

Number of Threads (parallel sessions executed)
Number of Seconds to run

You can then click on the Start button to start the test.

WebSurge Authorization required for non-local Sites

In order to run load tests against a remote server, you have to explicitly give access to it, by placing a websurge-allow.txt or robots.txt file in the root path of the server, or test against a local server. More info here.

The Request Progress Status Bar

While the test is running a Test Progress status is shown on the bottom of the WebSurge window.

This progress bar updates every second with the latest request statistics displaying the number of requests run, current request per second, how many requests have failed and more.

While the test is running the user interface remains active and you can continue to use WebSurge as usual. You can create and edit requests and fire individual requests. For best performance of the test in high load scenarios however we'd recommend just letting the test run without explicit UI interaction.

Test Summary Report

When the test is complete WebSurge displays a Test Result Report on the right, and a complete request list on the left.

WebSurge Test Result Report View

The report provides summarized rest result information which includes:

Total Requests run
Failed Requests
Requests per second
Average Request Time
Min and Max Request times
95th and 99th Percentile
1st and 5th Percentile
Total number of bytes sent and received
A per URL summary list

Printing the Test Summary Report

The result view is an HTML document and as such you can print the report by pressing Ctrl-P. This brings up the browser print dialog from which you can:

Print to a Printer
Print to PDF

Result Export

You can also export results to a variety of formats. There are both summary result exports that export just the summary data, as well as an export of all the test data. The full data export can be very slow as a lot of data may be exported if you ran long requests.