AX86U and AT&T 1 gig upload speed problem

[road hazard]

What do you mean by "put it in access point mode". If you're doing that then you're not touching any of the AX86U's routing functions or things like hardware acceleration, AiProtection, QoS, etc. So assuming you're using Ethernet connections and not WiFi you are effectively just using the LAN ports on the AX86U as a dumb switch.

In a thread that I referenced earlier, somebody did that and it fixed their problem. All of my testing is being done while hard wired directly into the AX86U. Factory reset with it in normal (router) mode and real world uploads are half of what they should be. Using it in AP mode (dumb switch) and same result after another factory reset.

 

[ColinTaylor]

Using it in AP mode (dumb switch) and same result after another factory reset.

Then something very strange is happening as the traffic is just being switched and not going "through" the router at all. It's just LAN to LAN traffic. I've just been testing that (iperf LAN to LAN) on my RT-AX86U and as expected the throughput in either direction is about 940 to 950 Mbps.

 

[road hazard]

Then something very strange is happening as the traffic is just being switched and not going "through" the router at all. It's just LAN to LAN traffic. I've just been testing that (iperf LAN to LAN) on my RT-AX86U and as expected the throughput in either direction is about 940 to 950 Mbps.

Not sure why I didn't try that earlier (LAN to LAN iPerf) ....... but this is giving me anywhere from 500-730Mbps with both systems hardwired. Maybe I do indeed have a defective router?! (During testing AIProtection, QoS and Traffic Analysis were all turned off.

 

[ColinTaylor]

Not sure why I didn't try that earlier (LAN to LAN iPerf) ....... but this is giving me anywhere from 500-730Mbps with both systems hardwired. Maybe I do indeed have a defective router?! (During testing AIProtection, QoS and Traffic Analysis were all turned off.

Clients using Jumbo frames or strange MTU settings?

 

[road hazard]

Clients using Jumbo frames or strange MTU settings?

Negative on both.

 

[ColinTaylor]

Negative on both.

And no other devices connected to it, wired or wireless, during the test?

 

[Tech9]

but this is giving me anywhere from 500-730Mbps with both systems hardwired

Are you testing Gigabit ports only or 2.5GbE port to Gigabit ports? This 2.5GbE port is on a separate controller.

 

[road hazard]

And no other devices connected to it, wired or wireless, during the test?

At the time, no. I re-did the test just now, my kids were rather upset ....... but wireless was disabled and nothing plugged in (even tried disconnecting the ONT) and it made no difference.

 

[road hazard]

Are you testing Gigabit ports only or 2.5GbE port to Gigabit ports? This 2.5GbE port is on a separate controller.

Wasn't using the 2.5Gbe port. Both test units were plugged into the normal LAN ports.

 

[ColinTaylor]

At the time, no. I re-did the test just now, my kids were rather upset ....... but wireless was disabled and nothing plugged in (even tried disconnecting the ONT) and it made no difference.

Bizarre. Can you test with different devices running the iperf client/server?

I've found generally that iperf can give inconsistent results (compared to other types of data transfer tests), particularly with certain versions of iperf and also with the Windows builds.

 

[road hazard]

Bizarre. Can you test with different devices running the iperf client/server?

I've found generally that iperf can give inconsistent results (compared to other types of data transfer tests), particularly with certain versions of iperf and also with the Windows builds.

If I get some spare time, and I'm really bored , I'll try different machines. I'm at the point now where I just want to throw my hands up. I've been banging my head against the wall the past few days trying to figure this out and it's getting frustrating. I'm convinced it's the router.

 

[road hazard]

Those settings are not managed directly by the user, but enabled/disabled by various features, like Ai Protection for example will disable h/w acceleration which is required for max throughout.

I had posted some useful links here in this thread about how to check if acceleration is disabled.

https://www.snbforums.com/threads/ax86u-and-at-t-1-gig-upload-speed-problem.79904/#post-778379

**If your properly reset the router then acceleration should be enabled. That would indeed be interesting if for whatever reason you had acceleration disabled, but yet could achieve ~900Mbps with a multi-stream iperf3 test.

Flow Timer Interval = 10000 millisecs
    Pkt-HW Activate Deferral rate = 1
    Pkt-HW Idle Deactivate = 0
    Pkt-SW Activate Deferral count = 0
    Flow Low Pkt Rate = 10
    Acceleration Mode: <L2 & L3>
    MCast Learning <Disabled>
    MCast Acceleration IPv4<Enabled> IPv6<Enabled>
    IPv6 Learning <Enabled>
    GRE Learning <Enabled>
    4o6 Fragmentation <Enabled>
    TCP Ack Prioritization <Enabled>
    HW Acceleration <Enabled>
    Notify Processing Mode <Hybrid>
    OVS Flow Learning <Disabled>
    Flow Learning Enabled : Max<16383>, Active<159>, Cummulative [ 293 - 134 ]

I ssh'd into my router and this is how things stand at the moment.

 

[road hazard]

I think I fixed it? In the most bizarre way imaginable though...........

To my friend:

------------------------------------------------------------
Client connecting to x.x.x.x, TCP port 5001
TCP window size: 85.0 KByte (default)
------------------------------------------------------------
[  3] local x.x.x.x port 54024 connected with x.x.x.x port 5001
[ ID] Interval       Transfer     Bandwidth
[  3] 0.0000-10.0047 sec  1.08 GBytes   927 Mbits/sec

From LAN to LAN:

------------------------------------------------------------
Client connecting to x.x.x.x, TCP port 5001
TCP window size: 85.0 KByte (default)
------------------------------------------------------------
[  3] local x.x.x.x port 54778 connected with x.x.x.x port 5001
[ ID] Interval       Transfer     Bandwidth
[  3] 0.0000-10.0012 sec  1.10 GBytes   942 Mbits/sec

The fix....... I remember reading a LONG time ago about how using USB 3.0 drives on routers could -sometimes- slow down wifi. People would recommend switching the USB mode to 2.0 to fix this. I don't have a USB drive attached to my router so I figured, eh..... I'm not having a wifi problem but I'll go ahead and switch it to USB 2.0 mode and low and behold..... 930Mbps speeds across the board on upload!

 

[ColinTaylor]

The fix....... I remember reading a LONG time ago about how using USB 3.0 drives on routers could -sometimes- slow down wifi. People would recommend switching the USB mode to 2.0 to fix this. I don't have a USB drive attached to my router so I figured, eh..... I'm not having a wifi problem but I'll go ahead and switch it to USB 2.0 mode and low and behold..... 930Mbps speeds across the board on upload!

Does the problem reappear if you switch back to USB 3.0? I've just tried switching between USB 2.0 and 3.0 and it made no difference to my iperf tests.

 

[SAL9K]

Wow, interesting, and good find! I can confirm similar behavior on WLAN<->LAN iperf3 to iphone 13 mini.

No USB drive, set to USB 3.0: ~850Mbps
No USB drive, set to USB. 2.0: ~850Mbps
w/ USB 3.0 drive, set to USB 3.0: ~600Mbps
w/ USB 3.0 drive, set to USB 2.0: ~850Mbps

However, w/out a USB drive connected, it doesn't matter what the USB setting is, I still get expected/full throughput. Maybe someone can shed light as to what is happening here. I don't use the ax86u as a file server, but knowing this is a problem is tickling my OCD fibers in all the wrong ways, lol!

 

[SAL9K]

Does the problem reappear if you switch back to USB 3.0? I've just tried switching between USB 2.0 and 3.0 and it made no difference to my iperf tests.

This agrees with my testing too; I needed to have a physical USB drive attached to effect a difference with the router USB setting.

 

[ColinTaylor]

This agrees with my testing too; I needed to have a physical USB drive attached to effect a difference with the router USB setting.

You're talking about a WiFi interference issue. I'm talking about Ethernet to Ethernet connections. I found no difference between USB 2.0 or 3.0, with or without a USB drive plugged in. The OP said he doesn't have a USB drive plugged into his router.

 

[road hazard]

Does the problem reappear if you switch back to USB 3.0? I've just tried switching between USB 2.0 and 3.0 and it made no difference to my iperf tests.

For me, yes.... if I switch it back to USB 3.0 mode, the upload speed gets cut in half so it's 100% reproducible.

 

[SAL9K]

Ok, it did make a difference with me; WLAN<->LAN throughput went down if a drive is connected and the USB setting is 3.0.

I think @road hazard is reporting that they didn't have a drive connected and the USB 2.0/3.0 setting makes a difference. I needed an attached drive to make a difference... which is a difference.

 

[road hazard]

Ok, it did make a difference with me; WLAN<->LAN throughput went down if a drive is connected and the USB setting is 3.0.

I've never had a drive plugged into this router (and have no need too) so leaving it in USB 2.0 mode isn't an issue. Just weird that for me, it made a clear difference and is reproducible.