RAXE300 Inconsistent Upload Speeds with Armor Enabled
I replaced my trusty R7800 with an RAXE300 one month ago. The RAXE300 is running firmware version 1.0.8.64_3. I have been pleased with the performance, range, and stability of the RAXE300, but I have experienced inconsistent Speedtest.net app upload results that I never saw with the R7800. The RAXE300 came with one-year of Netgear Armor, which I enabled.
I have a symmetric 1 Gbps fiber connection from Frontier. Testing from a wired Windows 11 machine using the Speedtest.net app to a Frontier Speedtest server 150 miles away, I have been experiencing inconsistent Internet upload speeds. Between 20% to 40% of Speedtests the last month have measured Internet upload speeds as low as 135 Mbps. The download speeds have consistently been greater than 850 Mbps, but typically the full 940 Mbps. I never experienced low upload speeds with my R7800. In contrast, if I use the Speedtest.net web-based test, or any of the other web-based speed tests, I never see low upload speeds.
Yesterday, I disabled Netgear Armor. In 24 hours of testing, all of the upload Speedtest.net results have been greater than 900 Mbps. My conclusion is that Armor is responsible for the slow upload speed tests. Both the Speedtest.net app and the web-based Speedtest use four TCP/IP connections with a large TCP window size. It would appear that Armor treats the web TCP packets differently than other TCP/IP connections. Has anyone else experienced this behavior?
In retrospect, I am not surprised that Armor might not be able to handle Gbps upload speeds since symmetric Internet connections at a Gigabit or higher have only become common in the last few years. In 2018 when Armor was first introduced, there were many complaints of low throughput. I think Netgear should be letting its customers know how Armor affects throughput.
