Are you running SpeedTest via Cato Cloud and seeing degraded results? No worries. Understanding the SpeedTest technique and the impact of Cato Cloud topology on the test result is crucial to get the most out of the test.
Running SpeedTest out of Cato Cloud network, will automatically force the selected SpeedTest server to be closest to connected Cato PoP. Having said that, without even starting the test, additional hop was just added to the equation, selecting non-optimal SpeedTest server resulting lower throughput comparing to direct connection to the internet.
For more information please refer to the SpeedTest documentation.
At last, Cato's team highly recommend to stick with same server each test (so no other server will be picked during latency check. You can read more how to do it, here)
Finally, running speed tests through a security proxy such as Cato can often give results that do not reflect actual performance of real user traffic. Speed test sites generate random data that appears as new or unique content. This triggers deeper security scanning than would be experienced with the vast majority of files on the web. The best way to test actual performance would be to download a well known, large file from a well connected host, and timing it manually. Suggestions could include ISO images of your favorite Linux distro, or a standalone service pack from Microsoft.
FYI - The Socket WebUI (Tools tab) offers a speed test tool that will run directly from the socket and it has the option to bypass Cato cloud in order to compare the results. See Using the Socket WebUI to Troubleshoot Links.