|
|
|
|
|
by med00d
4163 days ago
|
|
The accuracy can vary, but for the most part I've found them to display results that are right in line with what ISPs quote their speeds to be. Ookla explains the way that their speed tests work here -https://support.speedtest.net/hc/en-us/articles/203845400-Ho... and that's the idea behind all of them. Probably the most inaccurate representation of speed that I've come across was when I had just done a firewall replacement for a customer and I didn't catch that their ISP had the speed/duplex settings on their equipment set to a hard-coded 100/Full-duplex. The firewall that I installed was set to auto-negotiate and since the ISP's equipment was hard-coded, the firewall negotiated at 100/Half-duplex. If you're not familiar with what half-duplex means, basically it means that you can still send & receive data, but you can only do one at a time -vs- full-duplex where you can send and receive at the same time. The problem with a speed test is that the download test only downloads, and the upload test only uploads, so in the case of the duplex mismatch, the speed test looked fine. As soon as users came in the next morning and put the internet connection under load, they began complaining about internet speeds and we realized that we had a problem with the duplex settings. Another way to test is a utility called iperf, but it's a lot more complicated and requires you to have access to the remote machine. - https://iperf.fr/ |
|