Solved GT-AX6000 2.5G LAN Port - only negotiating at 1G on switch

[visortgw]

Well, if I do - it will likely be a PCIe card that I'll stuff in the case. Not a fan of those usb-c dongles to be honest.

Those look to be super cheap: https://www.amazon.ca/TP-Link-2-5GB-PCIe-Network-TX201/dp/B0BG685PKM/?tag=smallncom-20

But the dongles work great for a laptop or NAS...

 

[shabbs]

But the dongles work great for a laptop or NAS...

Absolutely! I've got a full sized case for my desktop machine down here so makes sense to use a PCIe card inside the case.

 

[jksmurf]

@visortgw quick Q pls, my TL-SG105E LEDs can be disabled in the Switch Web Admin, can the TL-SG108-M2 Switch LEDs also be disabled the same way?

 

[visortgw]

@visortgw quick Q pls, my TL-SG105E LEDs can be disabled in the Switch Web Admin, can the TL-SG108-M2 Switch LEDs also be disabled the same way?

No, there is no "web admin" since the TL-SG108-M2 is an unmanaged switch.

 

[jksmurf]

No, there is no "web admin" since the TL-SG108-M2 is an unmanaged switch.

Ah; ok thanks. So back to the black nailpolish then. ta

 

[maxbraketorque]

Dredging up this thread. My GT-AX6000 2.5g LAN port is acting like its in 1g mode despite the router log showing it in 2.5g mode. When I run the router WAN speed test, I get 2 gbps DL speeds, but when I run speed test from the Speedtest.net app on my computer with it directly attached to the 2.5g LAN port, DL speeds top out at 1 gbps. I'm using the same remote server for the speed tests on the router and the computer. The ethernet adapter on the computer is in 2.5g mode, and when I do file transfers across my network, I see full 2.5 g speeds in both directions. I'm using a Cat 6 cable. Everything is plugged into the correct port.

Any thoughts on the difference in internet DL speed between the router and the computer?

 

[degrub]

What does that remote speed test give you when you plug the PC lan port, using the same cable that you connected to the router with, and run the test ?

Since the router only has 1 port for WAN/LAN 2. 5 Gb/s and one LAN port at 2.5 Gb/s with the other 4 at 1 Gbit/s , how are you getting full 2.5 Gbit/s speeds in both direction for file transfers across your network ?

Might help to sketch the topology for your LAN speed test doing file transfers - BTW, use iperf to do the LAN testing, not a file transfer. Fewer variables .

 

[maxbraketorque]

I have yet not plugged my LAN PC directly into my cable modem. There was no time for that today. I'll try to do that in the morning.

Sorry for not fully describing the topology. The router 2.5g LAN port is normally plugged into a Netgear multigig managed switch that creates the hub of my wired LAN network. My AP (another GT-AX6000), my desktop PC, and my NAS are also plugged into this. LAN speed tests were between the desktop PC and the NAS (2.5g network port).

WAN speed tests gave the same result (1 gbps) whether my PC was plugged into the switch or directly plugged into the router 2.5g LAN port.

 

[Ripshod]

@maxbraketorque you haven't mentioned it yet as far as I can see, but what is the speed of the LAN port on your pc? Speeds have to be supported at both ends and along a link.

 

[Tech9]

what is the speed of the LAN port on your pc?

See post #46

The ethernet adapter on the computer is in 2.5g mode, and when I do file transfers across my network, I see full 2.5 g speeds in both directions.

 

[degrub]

I have yet not plugged my LAN PC directly into my cable modem. There was no time for that today. I'll try to do that in the morning.

Sorry for not fully describing the topology. The router 2.5g LAN port is normally plugged into a Netgear multigig managed switch that creates the hub of my wired LAN network. My AP (another GT-AX6000), my desktop PC, and my NAS are also plugged into this. LAN speed tests were between the desktop PC and the NAS (2.5g network port).

WAN speed tests gave the same result (1 gbps) whether my PC was plugged into the switch or directly plugged into the router 2.5g LAN port.

Netgear switch model number, please ?

 

[Tech9]

I don't think we need any model numbers. Based on issue description: all 2.5GbE clients showing the same lower WAN-LAN throughput -> it's obviously the router; specific 2.5GbE client only affected -> it's obviously the client or connection. I wouldn't trust the built-in speed test and we know nothing about router's configuration. I would agree it's a bug or some incompatibility only if the router was running factory firmware in default configuration with no Trend Micro components involved. The original issue this thread was created about is unrelated - it was just connections error.

 

[maxbraketorque]

I plugged the PC directly into the modem, and after a modem reboot to get the modem talking directly to the PC, the PC got 1850 mbps down on SpeedTest app, so the router is definitely the issue for me.

To answer the question about router config - its pretty basic. No Trend Micro. One guest network. No Merlin FW add-ons. Perhaps the only thing that's not basic is having a site-to-site VPN tunnel (using OpenVPN). Disabling the VPN tunnel had no effect. Rebooting the router had no effect.

I'm open to suggestions any router settings to change.

 

[Ripshod]

Swap cables, see if anything changes.

 

[maxbraketorque]

I've tried different cables. I should have mentioned that in my first post in this thread.

 

[degrub]

Factory reset. install ASUS std firmware. Factory reset again. Manual configuration only minimums. This will let you start from a known good configuration.

 

[maxbraketorque]

Factory reset. install ASUS std firmware. Factory reset again. Manual configuration only minimums. This will let you start from a known good configuration.

That's outside of what I'm willing to do, especially considering that I've never had any complex configs on this router. I'm willing to "suffer" with 1 gbps internet DL speeds.

If this is a real issue with the LAN port performance, I figure that other people will report it as well. The challenge here is that its hard to justify complaining about not getting more than 1 gbps DL internet performance, so most people with 2 gbps service may never notice or may never broadcast the issue. Its a very "1st world" issue.

 

[Tech9]

I'm open to suggestions any router settings to change.

Seems like you know what are you doing and you have covered the basics already. My only comment is on "No Trend Micro". This means you don't use any of the firmware components relying on bwdpi engine, you have withdrawn the data sharing agreement and have rebooted the router after. If all true - start fresh is perhaps the only thing left to try, but I also wouldn't bother.

 

[shabbs]

Dredging up my old screw up. LOL. Sigh... I feel your pain @maxbraketorque. This is going to haunt you until you can get to root cause.

I was curious and did a test just now from my wired Win11 Desktop using Speedtest.net as well. Got full speed results. Desktop is running a 2.5G PCIe card now. Connection goes to the 2.5G Switch which is connected to the 2.5G LAN port on the GT-AX6000. Very odd observation you have. Does it hold the same with any speed test site? What about https://fast.com/ or https://speed.cloudflare.com/ sites? A factory reset may be your only recourse at this point. It would not be too bad - save your current config so you can get back to it after with the same FW version.

My results just now:

 

[maxbraketorque]

Seems like you know what are you doing and you have covered the basics already. My only comment is on "No Trend Micro". This means you don't use any of the firmware components relying on bwdpi engine, you have withdrawn the data sharing agreement and have rebooted the router after. If all true - start fresh is perhaps the only thing left to try, but I also wouldn't bother.

You hit it the nail on the head. I had all Trend Micro features turned off, but I had not withdrawn. After I withdrew from the agreement, the speedtest result on my PC shot up to 2 gbps on the DL. ugh. Kinda interesting that a built-in feature prevents another built-in feature on the router from working. Does the BE86U or BE88U do any better?