What's new

Some kind of 2.5gbit bottleneck issue - testing new GTAX11000 on Merlin latest 388.7

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Swistheater

Very Senior Member
So I have just purchased a new GT-AX11000. I installed the latest stable firmware. I completely factory reset. I am gradually configuring the router. I have a 2.5gbit wan setup. I am plugging 1gbit devices into the 1gbit ports. For some reason with only hardware acceleration enabled, I am lucky to get between 400mbps and 500mbps download with wan connected to 2.5gbit port and 1gbit devices plugged into 1gbit ports. If I switch the wan connection to the 1gbit wan port, my speed tests range rised to be between 800 and 900 mbps. Is this normal bottleneck behaviors? The router itself is showing the connection is rated at 2.5gbits when connected on the 2.5gbit port, and the network cable is cat6. I also have the same issue when testing with a cat5e cable. What am I missing here?
 
So to understand this correctly your using the 2.5g port for wan internet.

-Is the isp modem (fibre,vdsl,cable,etc) capable of 2.5g as well? Without that your limited to 1g regardless of the port you use on the gt-ax11000.

-Have you tried any iperf tests from device to device? 2.5g iperf on router -> 2.5g iperf on pc?

-Hardware acceleration you said is on. Is runner (archer) and flow cache enabled? And are you using any QoS?

-Have you attempted to use ethtool or the port map in ui to verify the wan 2.5g port is actually connecting at 2.5g speeds and not degrading itself to a 1g connection.

Other miscellaneous things to check is your mtu set correctly? Ethernet automatic dhcp is generally 1500 mtu or 1492 for pppoe connections. When you ping is their fragmentation happening.

What device are you testing speeds on? The router?, a wired pc?, a wireless pc? And what speedtest are you using? Speedtest.net?

My own personal remarks as I own a gt-ax11000 non pro. The 2.5g seems to be fine for me. However I don’t use it as wan instead as lan for my aimesh nodes so there might be a difference there. It’s hard to test lan speed on that port except for direct connection using iperf as the router only has 1 2.5g port their would be not equivalent switching, but it shouldn’t drop speed well below what your getting on 1g ports.
 
So I have just purchased a new GT-AX11000

Can you send it back and get GT-AX11000 Pro instead?

GT-AX11000 is >5 years old model (2018) based on older BCM4908 CPU + BCM43684 radios hardware. It will never get Asuswrt 5.0 (3.0.0.6) and the support will be much shorter. The updated Pro version (2022) is based on BCM4912 CPU + BCM6715 radios hardware and has Asuswrt 5.0 available.
 
So to understand this correctly your using the 2.5g port for wan internet.

-Is the isp modem (fibre,vdsl,cable,etc) capable of 2.5g as well? Without that your limited to 1g regardless of the port you use on the gt-ax11000.

-Have you tried any iperf tests from device to device? 2.5g iperf on router -> 2.5g iperf on pc?

-Hardware acceleration you said is on. Is runner (archer) and flow cache enabled? And are you using any QoS?

-Have you attempted to use ethtool or the port map in ui to verify the wan 2.5g port is actually connecting at 2.5g speeds and not degrading itself to a 1g connection.

Other miscellaneous things to check is your mtu set correctly? Ethernet automatic dhcp is generally 1500 mtu or 1492 for pppoe connections. When you ping is their fragmentation happening.

What device are you testing speeds on? The router?, a wired pc?, a wireless pc? And what speedtest are you using? Speedtest.net?

My own personal remarks as I own a gt-ax11000 non pro. The 2.5g seems to be fine for me. However I don’t use it as wan instead as lan for my aimesh nodes so there might be a difference there. It’s hard to test lan speed on that port except for direct connection using iperf as the router only has 1 2.5g port their would be not equivalent switching, but it shouldn’t drop speed well below what your getting on 1g ports.
The ISP is comcast so the default MTU should be fine. The issue I am having is strictly a connection speed issue with the 2.5g port. All the 1g ports are working fine. When I do a speed test with the wan cable connected to various wan port locations, here is what I get.

Wan cable connected to 2.5gbit port, Testing Client connected to port 1gbit port on the routers switch board:

1720990377235.png


Wan cable connected to 1gbit port, Testing Client connected to port 1gbit port on the routers switch board:

1720990570965.png


What I am trying to figure out is why is there such a huge discrepancy? ethtool says the connection is 2.5gbit.
 
Can you send it back and get GT-AX11000 Pro instead?

GT-AX11000 is >5 years old model (2018) based on older BCM4908 CPU + BCM43684 radios hardware. It will never get Asuswrt 5.0 (3.0.0.6) and the support will be much shorter. The updated Pro version (2022) is based on BCM4912 CPU + BCM6715 radios hardware and has Asuswrt 5.0 available.

A bit premature to say it will never get 3.0.0.6 when supported models besides the initial haven’t been decided or publicly announced by rmerlin. Maybe you have information the rest of us don’t. 🤨 Or maybe I’m being optimistic in the absence of factual information.
 
Last edited:
The ISP is comcast so the default MTU should be fine. The issue I am having is strictly a connection speed issue with the 2.5g port. All the 1g ports are working fine. When I do a speed test with the wan cable connected to various wan port locations, here is what I get.

Wan cable connected to 2.5gbit port, Testing Client connected to port 1gbit port on the routers switch board:

View attachment 60170

Wan cable connected to 1gbit port, Testing Client connected to port 1gbit port on the routers switch board:

View attachment 60171

What I am trying to figure out is why is there such a huge discrepancy? ethtool says the connection is 2.5gbit.

I won’t be able to test this today, but I’ll give iperf a shot and see what kinds of speeds I’m getting from the 2.5g port. Unfortunately I don’t have wan speed similar to give you an Apple to Apple comparison. Nothing off the top of my head really should be affecting it, certainly could just be a bad implementation of 2.5g by Asus either in software or hardware. Do note there was multiple revisions of this router so maybe that could be why.
 
I won’t be able to test this today, but I’ll give iperf a shot and see what kinds of speeds I’m getting from the 2.5g port. Unfortunately I don’t have wan speed similar to give you an Apple to Apple comparison. Nothing off the top of my head really should be affecting it, certainly could just be a bad implementation of 2.5g by Asus either in software or hardware. Do note there was multiple revisions of this router so maybe that could be why.
I am pretty confident it is a hardware issue with the 2.5g router port on this router I have, I am mainly seeking confirmation at this point. When You have time, feel free to test yours.
 
Can you send it back and get GT-AX11000 Pro instead?

GT-AX11000 is >5 years old model (2018) based on older BCM4908 CPU + BCM43684 radios hardware. It will never get Asuswrt 5.0 (3.0.0.6) and the support will be much shorter. The updated Pro version (2022) is based on BCM4912 CPU + BCM6715 radios hardware and has Asuswrt 5.0 available.
What is your opinion of the GT-AXE16000 and GT-BE98_PRO in comparison with the GT-AX11000 Pro? I may pick one of those up.
 
A bit premature to say it will never get 3.0.0.6 when supported models besides the initial haven’t been decided or publicly announced by rmerlin. Maybe you have information the rest of us don’t. 🤨 Or maybe I’m being optimistic in the absence of factual information.
yea, I was wondering the same. he seems pretty knowledgeable about the internals, maybe he has some insiders source on the firmware continuation as well. might be an asuswrt whistleblower.
 
A bit premature to say it will never get 3.0.0.6 when supported models besides the initial haven’t been decided or publicly announced by rmerlin. Maybe you have information the rest of us don’t. 🤨

Expect GT-AX11000 and similar BCM490x routers EoL in next 2 years with few official Asuswrt 4.0 firmware updates. RMerlin can't migrate any of them to 102 and may eventually continue for some time with 388 maintenance updates as usual. Happening right now to EoL AC-class routers on 386.

What is your opinion of the GT-AXE16000 and GT-BE98_PRO in comparison with the GT-AX11000 Pro? I may pick one of those up.

They'll have expected longer support at least. Newer hardware inside and Asuswrt 5.0 compatible. Good, bad, needed, etc. for your use case - I don't know. I see you have collected quite a few EoL Asus routers in your signature. Perhaps time to update some of them to extend the firmware support.
 
Last edited:
A bit premature to say it will never get 3.0.0.6 when supported models besides the initial haven’t been decided or publicly announced by rmerlin. Maybe you have information the rest of us don’t. 🤨 Or maybe I’m being optimistic in the absence of factual information.

non pro routers will not get the latest 3.0.0.6 5.0 FW been stated a few times , pro models might get the upgraded FW
 
non pro routers will not get the latest 3.0.0.6 5.0 FW been stated a few times , pro models might get the upgraded FW

That’s fine. My next routers will probably be a Dream Machine Pro Max and Altalabs AP’s. Depending on when my needs arise.

Better to get the toys I want then fiddle with things that can’t or probably shouldn’t. 10-15 years down the road I imagine Asus will be much the same hw and sw wise.
 
A bit premature to say it will never get 3.0.0.6 when supported models besides the initial haven’t been decided or publicly announced by rmerlin. Maybe you have information the rest of us don’t. 🤨 Or maybe I’m being optimistic in the absence of factual information.
You'll have to forgive blanket statements like that from those of us who have been around long enough to be losing our stuffing: we often forget we have to cite our evidence/reason for being so certain
 
especially for gt-axe16000 and gt-axe11000 i find it hard to believe they will not get the next version and according to your terminology the are "non pro"
 
You'll have to forgive blanket statements like that from those of us who have been around long enough to be losing our stuffing: we often forget we have to cite our evidence/reason for being so certain

That's fine just thought I missed something explicitly said by rmerlin. I don't doubt the logic.
 
Screenshot 2024-07-15 170014.png


Router to pc; Router 2.5g to 10g nic set to 2.5g on pc.

Hw acceleration is disabled for me... Just remembered since I use cake.
 
Last edited:
Screenshot 2024-07-15 171204.png


No real difference interestingly even worse with qos cake off.
 
Last edited:
What is "Router to PC"?

Also NAT acceleration and QoS are WAN-LAN, nothing to do with LAN-LAN.
 
What is "Router to PC"?

Also NAT acceleration and QoS are WAN-LAN, nothing to do with LAN-LAN.

I just meant a straight Lan to Lan connection without any switches or degrading of links speeds. Since the router is using the 2.5g port to the shortest path.

I see, understood. Anyways as I said I cannot test an apple to apple comparison since I don’t have comparable wan speed. Since he is getting his wan speedtest with hardware acceleration enabled he has a higher rating then I’m getting with iperf. Im probably also fragmenting down to 1500mtu from 9000mtu testing directly to the router.

But I think it will still be somewhere in that ballpark.
 

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top