Did not seem to work, unfortunately. Also tried disabling tso
Yeah, I was afraid of that, I am noticing a high number of retries when using 2.5G speeds that I just cannot locate the reason for.
Edit : I wont bother doing another paste, but I can go right up to 2.3gb (-b 2300mbit) and the retries are in the low teens over a 10 second run.
Edit 2 : Maybe this has somthing to do with the fact that the "other" (non 301W) end does not support pause frames?
Supported pause frame use: No
As soon as I drop the rate to just below the theoretical max things calm down quite a bit:
root@HOSTNAME:~# timeout 5 iperf3 -c 192.168.1.1 ; timeout 5 iperf3 -c 192.168.1.1 ; timeout 5 iperf3 -c 192.168.1.1 -b 2250mbit ; Connecting to host 192.168.1.1, port 5201 [ 4] local 192.168.1.29 port 58654 connected to 192.168.1.1 port 5201 [ ID] Interval Transfer Bandwidth Retr Cwnd [ 4] 0.00-1.00 sec 283 MBytes 2.37 Gbits/sec 94 382 KBytes [ 4] 1.00-2.00 sec 280 MBytes 2.35 Gbits/sec 75 472 KBytes [ 4] 2.00-3.00 sec 279 MBytes 2.34 Gbits/sec 160 380 KBytes [ 4] 3.00-4.00 sec 278 MBytes 2.33 Gbits/sec 263 269 KBytes [ 4] 4.00-4.95 sec 264 MBytes 2.32 Gbits/sec 139 267 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bandwidth Retr [ 4] 0.00-4.95 sec 1.35 GBytes 2.34 Gbits/sec 731 sender [ 4] 0.00-4.95 sec 0.00 Bytes 0.00 bits/sec receiver iperf3: interrupt - the client has terminated Connecting to host 192.168.1.1, port 5201 [ 4] local 192.168.1.29 port 58674 connected to 192.168.1.1 port 5201 [ ID] Interval Transfer Bandwidth Retr Cwnd [ 4] 0.00-1.00 sec 242 MBytes 2.03 Gbits/sec 0 288 KBytes [ 4] 1.00-2.00 sec 268 MBytes 2.25 Gbits/sec 0 301 KBytes [ 4] 2.00-3.00 sec 267 MBytes 2.24 Gbits/sec 14 192 KBytes [ 4] 3.00-4.00 sec 264 MBytes 2.21 Gbits/sec 17 246 KBytes [ 4] 4.00-4.95 sec 255 MBytes 2.24 Gbits/sec 0 279 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bandwidth Retr [ 4] 0.00-4.95 sec 1.27 GBytes 2.19 Gbits/sec 31 sender [ 4] 0.00-4.95 sec 0.00 Bytes 0.00 bits/sec receiver iperf3: interrupt - the client has terminated root@RASPBERRYPI-01:~#
Well, basic issue is that network driver is shit when it comes to performance and sooner or later we are hitting the raw CPU performance limit
Great that you have 2 devices, your getting 2.3mbps (vs a max 1.9 to each in my set-up), which is more than i get! so give me hope and some new test options (based on your retries) i will also build an image to the NSS offload and post results soon
No one has an idea? Or at least can you narrow down if it is a hardware defect? I can still return or exchange the device until 2.4.2023.
Take a look at system log to see if you can find the cause.
Choose chan 36@80Mhz in the router to see if the cause was DFS.
PS: I don't have this router.
Thank you for your answer, what do you mean with DFS?
If radar is detected in the channel you are using, DFS will cause the radio to turn off that channel. Search about DFS to know more.
Ok, if them so, is there a possible to deactivate "DFS"?
Dont use DFS channels then, in the USA it is a requirement for wifi devices using DFS channels to listen before enabling tx and if they pick up a signal that may look like radar to turn off tx and listen. This is a legal requirement.
I don't understand anything anymore. I'm sorry. What is DFS and how can I disable it? I do nothing different as usual in my previous routers. I have set Mode AX, channel 36 and 160 MHz. Is there anything wrong with that?
in lay terms ... if a wifi is interfering with an airport radar for incoming planes ... the airport radar takes priority! which country you are using ? channel 36 / 160mhz is not a DFS channel i don't think in the US
I still have Netgear APs in use with the same WLAN configuration. On these devices the WLAN is not deactivated until now. Country is DE.
160Mhz wide channels always overlap with DFS required frequencies (besides the rather new U-NII 4 band, but that's only partially usable in Europe)
It's possible that your existing netgear router doesn't detect a radar signal, but if there is a radar detection, the wifi device needs to disable the radio temporarily, this is a regulatory requirement.
But the solution ist simple, don't use 160Mhz wide channels, the performance increase is rather small
I'm sure you did search the internet for DFS and found this article on Wikipedia (in German).
The question of @sqrwv still stands, please have a look at the system log to see what entries you find at the time regarding 5GHz Wifi, maybe it's DFS related but maybe not.
You choose the not-DFS channel 36@80Mhz as suggest and see if it still happens.
Ok as I understand now is that I am not using a DFS channel with 36(5180MHz)@160MHz.
The DFS Channels in DE are from 5,25–5,35 GHz and 5,47–5,725 GHz. So I have to wait until the next problem and then I have to check the syslog.
I also have no problem to go to 36@80Mhz to avoid errors. I just thought because many in the forum put great value on 160Mhz that it would be much better. As long as it's not a hardware defect, I don't care. Thanks all for your answers.
^It's because of the above, that "chan36@160Mhz" is using DFS.
5170Mhz to (5170+160=)5330Mhz
chan36 to chan64
bearing in mind the regdomain issues with QCA/Qualcomm if you can set it to US it willl help you overcome these issues until QCA comes out with an up to date regdb.bin it is beyond the control of anyone than QCA to fix their outdated regdomains as all these settings are on their firmware
What is the max power limit for this unit for Panama with QCA outdated regdomains, I'm curious.
What about the wireless coverage and signal strength compared to R7800 that I consider as standard for long range and high speed WLANs (both 2.4 and 5 GHz). Hardly to be beaten by any other router especially with longer distance.