Performance Testing Tool Selection For Desktop Based Application

There are 2 approaches and you need to choose which one is suitable for your performance testing activities:

1. If you need to load test desktop application itself - you should go for a profiler tool.

2. Classic performance testing tools simulate hundreds/thousands of virtual users so if you want to simulate multiple desktop applications concurrently accessing the backend or database - in this case it is possible to simulate network footprint of the desktop application and replay it with increased load.

   With regards to network protocols - the performance testing tool needs to support all network protocols used by your application. If you don't know what protocols are under the hood you can use a sniffer tool to analyze which are being used and then choose the performance testing tool which supports them.

   With regards to recording: some of performance testing tools allow recording and replaying, at least for HTTP protocol. You can take a look at How to Do Desktop Performance Testing article for more details.

If your desktop application communicates with a backend over standard protocols (HTTP/S, WebSockets, etc.), tools like JMeter or Gatling could be viable open-source options. However, if you need full protocol-level performance testing similar to what you did with LoadRunner’s RDP, your choices are more limited.

Since you’ve used LoadRunner before but are concerned about licensing costs, you might want to check out WebLOAD. It offers similar capabilities for complex performance testing but with a more cost-effective licensing model. It supports a wide range of protocols and provides correlation, scripting, and analytics features comparable to LoadRunner.

Would be helpful to know more about how your application interacts with the backend—does it use web protocols, or is it primarily an RDP-based system?