Gigabit Tx/Rx rates yet slow transfer speeds over wifi

[splendid_koi]

Certainly doesn't look normal to me. I have an RT-AX58U with 1.0 up/down Gbps fiber service from AT&T. Here is a speed test from an android phone. Granted this is an ideal condition (quiet network, AX connection, 5GHz, next room over from (actually above of) router.

View attachment 56382

This is much more in line with my expectations. I am closer to the router and also testing on AX, from 2 Apple and one HP clients.
I do live in a noisy environment though but on the other hand the router is a bit more capable and RSSIs, SNRs, and Noise levels are all excellent.

 

[glens]

Sounds like you've done your due diligence. Time to RMA the router?

 

[splendid_koi]

Here's another interesting tidbit: I installed Ookla CLI on the NUC. It tests at 112 down and 800-ish up (ONT > Asus > NUC all hardwired).
I also installed iPerf on my hardwired ATV4K and it tests at 950 down and 950 up to the iPerf server on the NUC (ATV > Asus > NUC all hardwired).
This is getting crazy now, makes no sense. I am going to go pickup a Netgear RS700 now for the fun of it and do some barebones testing with that.

 

[TheLostSwede]

You are aware that Ookla isn't a reliable test, right?
I get 70 Mbps upload speeds to one local server and 250-270 Mbps from most other local servers.
As such, you can't rely on Ookla, as you have to test towards multiple servers.
This is obviously not your main problem though, but it appears to be part of your issue.

 

[L&LD]

Why would the expectation be that your upload is a one fifth of your download, when you should have the exact same bandwidth and capability in each direction?

That is not my expectation.

But your testing methods are not consistent or examples of trying to get those maximum speeds, IMO. WiFi isn't a fixed commodity. It is a variable that has many factors affecting it. Including client device quirks and differences.

 

[IAAI]

https://www.snbforums.com/threads/asus-ax-axe-slow-upload-speeds.70352/

Fyi , even RT-BE96U suffers from the same symptoms

\\\\\\\

Don't waste your time looking for a solution.

asus and their reps will try really hard to blame it on your config, setup and environment. If you corner them they will only admit it qualifies for an RMA (so their devs can invistigate the issue), betting that you will get board communicating back and forth and then accepts the reality.

Asus AM5 motherboards

 

[splendid_koi]

You are aware that Ookla isn't a reliable test, right?
I get 70 Mbps upload speeds to one local server and 250-270 Mbps from most other local servers.
As such, you can't rely on Ookla, as you have to test towards multiple servers.
This is obviously not your main problem though, but it appears to be part of your issue.

Hence I tested locally with iPerf and got the same results?

 

[splendid_koi]

Netgear RS700 results in 700Mbps down and 400Mbps up on the 5GHz AX. Download is equal, upload is double that of GT-AX11000 (older hardware so ok I guess).
Netgear RS700 results in 1,400Mbps down and 800Mbps up on 6GHz AX (WiFi 6e). Download is equal, upload is 8x the upload of RT-BE96U on the same band.
Clearly, ASUS routers have issues with upload speeds over wifi, no matter the client. 5GHz band appears to be a bit better but the 6GHz band is a total distaster.
Considering the same issue is present even when ISP hardware is disconnected and I am testing to a hardwired iPerf server (on my NUC), the problem is 100% on ASUS.

 

[glens]

Clearly, ASUS routers have issues with upload speeds over wifi, no matter the client. 5GHz band appears to be a bit better but the 6GHz band is a total distaster.
Considering the same issue is present even when ISP hardware is disconnected and I am testing to a hardwired iPerf server (on my NUC), the problem is 100% on ASUS.

Neither myself nor many, many others have this issue, so MOST clearly it is NOT "an issue with Asus routers" in and of themselves. Either you've got a faulty unit or you're mis-configuring it in a way your other units mis-behave-to-suit when also so configured.

 

[splendid_koi]

Neither myself nor many, many others have this issue, so MOST clearly it is NOT "an issue with Asus routers" in and of themselves. Either you've got a faulty unit or you're mis-configuring it in a way your other units mis-behave-to-suit when also so configured.

A simple search across this very forum alone will show this to be untrue - quite a few others have the exact same issue, and I quoted a few before.
A faulty unit would be a good possibility however what are the chances that I got 3 faulty ASUS units, as I did test with brand new ones too?
Sure, misconfiguration could be an option but that wouldn't really be the case as I tried my old config, factory config, and another suggested config I found here.
There is also of course an option which you for some reason aren't willing to admit exists: that there is in fact an issue with ASUS units that has been unresolved.

 

[glens]

While I accept the possibility there's an Asus problem "in general", a very few cases out of many many thousands doesn't seem to indicate such to be the case. I have "played around with" the settings a great deal over time and this issue in particular sure seems to have ever gone un-triggered for me. But then I've never tried anything I know to be either useless (like maybe crippling an AX broadcaster to AC, etc), or even ill-advised. Perhaps I just have an innate knack...

The only times I recall having performance issues have been when using AiMesh, but that has always been with hardware making up the mesh being exactly the same throughout. I can only imagine what may befall a user trying to "mesh" disparate hardware.

I cannot at this time recall you talking about mesh, one way or the other. That could be a whole 'nother "can of worms". Do you encounter this problem while "meshed" or while not? And if meshed, is it wired or wireless, and/or does any node have different makeup (tri- or quad-radio meshing with dual-radio, etc.)?

 

[TheLostSwede]

Netgear RS700 results in 700Mbps down and 400Mbps up on the 5GHz AX. Download is equal, upload is double that of GT-AX11000 (older hardware so ok I guess).
Netgear RS700 results in 1,400Mbps down and 800Mbps up on 6GHz AX (WiFi 6e). Download is equal, upload is 8x the upload of RT-BE96U on the same band.
Clearly, ASUS routers have issues with upload speeds over wifi, no matter the client. 5GHz band appears to be a bit better but the 6GHz band is a total distaster.
Considering the same issue is present even when ISP hardware is disconnected and I am testing to a hardwired iPerf server (on my NUC), the problem is 100% on ASUS.

You still don't have a 6 GHz band on the GT-AX11000...

Below are the test results running iperf3 between my laptop and my desktop, using a GT-AX6000. It averages similar numbers as the Netgear RS700 you tested with. However, for some reason, the router doesn't appear to like my phone, despite synching at 2401 Mbps, I don't see speeds over 230-240 Mbps on the phone.

This leads me to a question here, have you tried with a few different clients? As different WiFi clients perform differently with different router hardware. Your router has Broadcom chips in it and would most likely work best with other Broadcom based devices, but as you can see, it has no issues with my Intel/Killer network card in my laptop, but doesn't like whatever module Google uses in the Pixel 6. That said, running Ookla's speed test, I get as I mentioned 250-270 Mbps on my phone...

Accepted connection from 192.168.50.56, port 53440
[  5] local 192.168.50.75 port 5201 connected to 192.168.50.56 port 53441
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-1.00   sec  46.3 MBytes   388 Mbits/sec
[  5]   1.00-2.00   sec  49.0 MBytes   411 Mbits/sec
[  5]   2.00-3.00   sec  45.4 MBytes   381 Mbits/sec
[  5]   3.00-4.00   sec  52.0 MBytes   437 Mbits/sec
[  5]   4.00-5.00   sec  52.3 MBytes   439 Mbits/sec
[  5]   5.00-6.00   sec  53.7 MBytes   450 Mbits/sec
[  5]   6.00-7.00   sec  47.4 MBytes   398 Mbits/sec
[  5]   7.00-8.00   sec  49.1 MBytes   411 Mbits/sec
[  5]   8.00-9.00   sec  49.8 MBytes   418 Mbits/sec
[  5]   9.00-10.00  sec  51.4 MBytes   430 Mbits/sec
[  5]  10.00-10.05  sec  2.62 MBytes   467 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-10.05  sec  0.00 Bytes  0.00 bits/sec                  sender
[  5]   0.00-10.05  sec   499 MBytes   417 Mbits/sec                  receiver

Also, by running -P 2, you can test with two parallel streams in iperf3, which yields higher speeds on my phone.

- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[  5]   0.00-10.04  sec   209 MBytes   174 Mbits/sec                  receiver
[  7]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[  7]   0.00-10.04  sec   208 MBytes   174 Mbits/sec                  receiver
[SUM]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[SUM]   0.00-10.04  sec   417 MBytes   348 Mbits/sec                  receiver

and laptop.

- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[  5]   0.00-10.04  sec   359 MBytes   300 Mbits/sec                  receiver
[  7]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[  7]   0.00-10.04  sec   358 MBytes   299 Mbits/sec                  receiver
[SUM]   0.00-10.04  sec  0.00 Bytes  0.00 bits/sec                  sender
[SUM]   0.00-10.04  sec   717 MBytes   599 Mbits/sec                  receiver

In other words, it helps to know how to use the various performance tests as well.
You can also try -P 4, which in the case of my phone yielded speeds of over 660 Mbps on average.

For a laugh, I ran -P 12 and got over 1 Gbps...

- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[  5]   0.00-10.02  sec   110 MBytes  91.8 Mbits/sec                  receiver
[  7]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[  7]   0.00-10.02  sec   113 MBytes  94.9 Mbits/sec                  receiver
[  9]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[  9]   0.00-10.02  sec   114 MBytes  95.3 Mbits/sec                  receiver
[ 11]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 11]   0.00-10.02  sec   110 MBytes  92.1 Mbits/sec                  receiver
[ 13]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 13]   0.00-10.02  sec   110 MBytes  91.9 Mbits/sec                  receiver
[ 15]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 15]   0.00-10.02  sec   112 MBytes  94.2 Mbits/sec                  receiver
[ 17]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 17]   0.00-10.02  sec   109 MBytes  91.1 Mbits/sec                  receiver
[ 19]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 19]   0.00-10.02  sec   108 MBytes  90.7 Mbits/sec                  receiver
[ 21]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 21]   0.00-10.02  sec   110 MBytes  92.1 Mbits/sec                  receiver
[ 23]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 23]   0.00-10.02  sec   109 MBytes  91.2 Mbits/sec                  receiver
[ 25]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 25]   0.00-10.02  sec   109 MBytes  91.6 Mbits/sec                  receiver
[ 27]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[ 27]   0.00-10.02  sec   109 MBytes  91.6 Mbits/sec                  receiver
[SUM]   0.00-10.02  sec  0.00 Bytes  0.00 bits/sec                  sender
[SUM]   0.00-10.02  sec  1.29 GBytes  1.11 Gbits/sec                  receiver

 

[doczenith1]

For a laugh, I ran -P 12 and got over 1 Gbps...

I have found that I can get similar speed results to using -P XX by using -w 4M without all the extra results syntax.

 

[TheLostSwede]

I have found that I can get similar speed results to using -P XX by using -w 4M without all the extra results syntax.

Nice one, I wasn't aware of that option. It resulted in even better numbers without running multiple parallel streams.

Accepted connection from 192.168.50.87, port 53150
[  5] local 192.168.50.75 port 5201 connected to 192.168.50.87 port 53158
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-1.00   sec   124 MBytes  1.04 Gbits/sec
[  5]   1.00-2.00   sec   176 MBytes  1.48 Gbits/sec
[  5]   2.00-3.00   sec   172 MBytes  1.44 Gbits/sec
[  5]   3.00-4.00   sec   166 MBytes  1.39 Gbits/sec
[  5]   4.00-5.00   sec   176 MBytes  1.48 Gbits/sec
[  5]   5.00-6.00   sec   176 MBytes  1.47 Gbits/sec
[  5]   6.00-7.00   sec   175 MBytes  1.46 Gbits/sec
[  5]   7.00-8.00   sec   152 MBytes  1.28 Gbits/sec
[  5]   8.00-9.00   sec   154 MBytes  1.29 Gbits/sec
[  5]   9.00-10.00  sec   171 MBytes  1.44 Gbits/sec
[  5]  10.00-10.05  sec  7.84 MBytes  1.29 Gbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  5]   0.00-10.05  sec  0.00 Bytes  0.00 bits/sec                  sender
[  5]   0.00-10.05  sec  1.61 GBytes  1.38 Gbits/sec                  receiver