A couple of weeks ago, we held our webinar “Beyond Protocol Results: Why and How to Analyze End User Experience through Real Browser-based Load,” in which we discussed modeling end user behavior correctly, increasing test coverage, running a load test using real browsers at scale, and more.
If you didn’t get a chance to attend, make sure you download the on-demand webinar recording!
Not surprisingly, there were a lot of questions about how LoadNinja differs from other load testing tools on the market that require more technical knowledge and effort. Here are many of the questions that were asked on the webinar. Hopefully they provide you with insights into how performance testing can be more accurate and less strenuous.
Can LoadNinja break down the response time of server side and browser side?
LoadNinja can report on the duration timings and the navigation timings to give an accurate metric that’s telling of the end-user timings, and how the network is able to handle the web application at scale. Read more about the LoadNinja navigation timings chart.
Navigation Timings chart
When you are looking at the results, do you also get the specific time it took for each transaction – not just graphs?
The Virtual User Debugger can be leveraged to analyze any errors that occurred during the lead test run. Additional details are also provided on the timings specified for each element that was loaded on the error-prone virtual user.
Can you show a data-driven demo, where the username is unique for each login?
Yes, absolutely. You can schedule a demo here.
In LoadNinja, testers can leverage the power of Data-Driven Testing by incorporating an external databank file (in either CSV or TXT format) and parameterize various inputs previously recorded in LoadNinja.
How does LoadNinja scale to 10 to 20 users for the data-driven test, and also apply this for a period of 5 minutes?
In order to scale a load test, you simply specify the amount of concurrent virtual users you wish to employ in a test through the Load Test Scenario page. There is no memory allocation or monitoring or local computing resources needed as LoadNinja’s servers will spin up the exact number of real web browsers for each concurrent virtual user designated in the load test configuration.
LoadNinja Scenario Page
Is there a way you can get dynamic value or values from the response, and use that value as a parameter in the following requests?
What should be the user to ramp up ratio?
It truly depends on the behavior corresponding with the website under test. If testers wish to model user activity where customers log in to a website all at once, then specifying 1 ramp up step will implement this modeled behavior.
Otherwise, increasing the ramp up step intervals will create a step function in which Virtual Users are released in time increments over the total course of the ramp up time specified.
Do we need to have JMeter running on the side to get protocol results between server and client or does that metric get recorded by LoadNinja?
Yes, this is where leveraging LoadNinja in concert with JMeter becomes advantageous. As having an accurate depiction of users’ experience generated from the UI layer while providing results associated with an open-source protocol-based tool can give detailed metrics telling of both sides of the application.
Can you show the HTTP protocol results in LoadNinja?
LoadNinja can provide the request, response, breakdown of the timings, and the HAR format report associated with each element on a page that throws an error during the load test run.
Are the metrics from the load generators available during run (i.e. CPU usage, memory, network)?
The Loadgen Health Tab in LoadNinja will show the metrics associated with the load generators used during the load test run. You can read more about it here.
Loadgen Health Tab
Does LoadNinja support performance testing PowerBI?
Yes, LoadNinja can capture recordings against a PowerBI web application just as an end user would interact with it through a browser window.
LoadRunner uses “correlation” to sync up encryption, dates, session IDs, and other dynamic data. How does LoadNinja accomplish this?
In a very similar manner. The main difference is that with LoadNinja, correlation is performed automatically and is completely abstracted from the user experience. This enables testers to create more accurate models that can be used in performance tests that are truly indicative of the user experience.
By recording and playing back interactions through a real web browser, LoadNinja natively supports the integration of all of these web technologies in frameworks. In the same manner that a web browser handles these items, LoadNinja captures their same usages in its novel load testing platform.
Ready to start load testing? Try our free, 14-day LoadNinja trial today.