cancel
Showing results for 
Search instead for 
Did you mean: 

Bad lag spikes on GT6 nodes every five minutes, noticeable when streaming games to steamdeck etc

Cloakedrat
Level 8

Hey all,

I recently replaced my old ASUS routers with four GT6s. In terms of speed they are hands down better than the old routers that I had been using. Unfortunately l've identified a significant issue and I'm not sure if it's unique to the GT6s or was always an issue with Aimesh which the better signal is now highlighting? Unfortunately I can't seem to fix it myself.

The issue

Every 5 minutes from the last start up of the wifi signal there is a large lag spike lasting 5-10 seconds. I've attached three examples on the 5ghz-1 channel using ping plotter attached. This happens on all three signals 2.4ghz, 5ghz-1 and 5ghz-2 but only impacts devices on an aimesh node, it does not happen to devices connected to the main router. This is disrupting game streaming through sunshine / moonlight, which is almost perfect otherwise, as well as online gaming and to a lesser extent video calls etc - it is not too noticeable on internet browsing or TV streaming as you would expect. The worst issue however is if you bound a device to a node and use the 5ghz-1 channel it can disrupt it enough to kick off the device from the network I assume as it attempts to connect elsewhere for 3-4 minutes every 5 minutes.

The set up

All GT6s are using the second 5ghz signal as backhaul, unfortunately Ethernet cable isn't practical for my set up and is largely why I got these devices. All GT6s are using the latest firmware (3.0.0.4.388_24580-gf5f687f) - they seemed to update on their own after a time even though auto update was off - ignoring that for now! Devices tested have been iphone, Apple TV and steamdeck. All channels are avoiding DFS.

Testing I have done to try to fix

Factory reset all nodes multiple times; tried using just one router and one node on a new SSID with only one device; moved the nodes to different areas; tried different channels, smaller bandwidths 40mhz only; turned off most of the wifi settings that were enabled and vice versa (eg. tried smart connect, turning of MIMO, beamforming etc). Tried plugging the steamdeck in to a node via Ethernet. All result in the same problem

Suspect problem

I noticed in some discussion on trying to fix similar issues people were having with game streaming on steamdeck it was suspected the wpa_supplicant’s background scan was causing the issue by reviewing wifi signals available exactly every 300 seconds. Using ssh to look at the aimesh nodes i have noticed they use wpa_supplicant but the main router does not. My guess is that maybe this is used by the nodes to connect to the backhaul and the bgscan default setting is running every 5 minutes disrupting it and causing the lag. While there are wpa_supplicant conf files you can see when using SSH they are all temporary and I don’t think you can restart the wifi without them being reset to try for a temporary fix?

So in the end I’m stuck, and trying to decide whether to return the GT6s if I can’t resolve this issue. Has anybody else had the same issue with GT6s or other aimesh nodes with wifi backhaul?

Thank you

 

11:05 issue pinging main router - 5GHZ-111:05 issue pinging main router - 5GHZ-1

 

11:10 issue pinging main router - 5GHZ-111:10 issue pinging main router - 5GHZ-1

 

11:15 issue pinging main router - 5GHZ-111:15 issue pinging main router - 5GHZ-1

 

848 Views
1 ACCEPTED SOLUTION

Accepted Solutions

Cloakedrat
Level 8

Hey all,

Hopefully my final post on this issue. 

Thanks for all the help, particularly Saltgrass. I would have sent the logs but they weren’t super useful because if no one was roaming around the house they would have nothing in them for hours despite the lag issue occurring every five minutes.

Anyhow, I originally didn’t think ASUS support were able to handle firmware issue requests but I was wrong. The RMA team were fantastic, didn’t ask me to send the devices back to them for inspection needlessly which I was concerned might happen, and quickly gave me a slightly modified version of the latest firmware that, when loaded on to the nodes, seems to have stopped the issue. Very appreciative of the quick turnaround here and action by ASUS support. Thank you

View solution in original post

17 REPLIES 17

Saltgrass
Level 14

You haven't seen anything in any logs showing a problem every 5 minutes?

If you look at the Network Map, check the processor utilization to see if one is spiking.

Is there anything in your environment that has a 5-minute cycle?

Maximus Z890 Hero,
Intel Core Ultra 9 285K

Cloakedrat
Level 8

Hey,

Thank you. Unfortunately nothing obvious.

The general logs have nothing in them over the last half an hour other than one roaming request and the wifi logs say interference is acceptable, no channels are dfs and noise is -90 - -100. I tried refreshing it during the last lag spike and nothing changed with the noise

The cpu usage doesn’t seem to spike either. I also used SSH to see the top processes on the nodes if there was anything obvious there but there wasn’t anything using a lot of processing, possibly some activity on conn_diag and maybe the wpa_supplicant processes. I killed all the conn_diag related processes and the checks that restart them on the nodes, and still had the issue unfortunately.

The exact timing of when the 5 minute cycle occurs changes a little apparently based on when the wifi or modem was last restarted so don’t think it’s environmental although I’m still thinking about it and thinking of some way to fully rule that out

IMG_0029.jpeg

You have one LAN port marked as yellow.  If you have an Apple TV, I know why.   But if you don't, do you know what is causing the problem?

Maximus Z890 Hero,
Intel Core Ultra 9 285K

Cloakedrat
Level 8

Probably similar to an Appletv it’s just a device in standby as when it’s turned on that port is green. I’ve tried removing all the other devices on the network, except one to test it, to see if that made a difference but no change sadly.

Wonder if anyone else with a 388 ASUS firmware/code base aimesh main router and node could see if they have similar lag spikes every five minutes when the testing device is connected or bound to the node to try to see if it is gt6 specific or not? I used PingPlotter on iOS to test it

 

Can you unplug or remove the power from the device? 

My Apple TV behaved the same way with the port.  What I discovered is I was using Ethernet and had a Wireless backup connection.  It turned out those two networks were fighting over which one was doing the streaming so, when it jumped, it stopped the streaming while that was happening.

Once I removed the Wi-Fi network on the Apple TV, my streaming stabilized.

If you check your log, you may see the router is trying to connect to that device even when it is in standby.

Do you feel the Ping Plotter software is safe?   Maybe the fact I can go for hours without my streaming having problems, indicates I do not have any such interruptions.

Maximus Z890 Hero,
Intel Core Ultra 9 285K

Cloakedrat
Level 8

Great idea as I do have a few devices that create their own network, particularly when they disconnect. There’s nothing in the logs that repeats with that level of frequency but I did see a node disconnect and reconnect at the time of one of the spikes which could just be a coincidence and confirmation bias that it’s the nodes checking their wifi connection every five minutes causing the issue.  I’ll try turning everything off and see if that makes a difference.

PingPlotter seems ok for iOS devices, but definitely could just use terminal or cmd prompt and review it manually.

Main things to have a somewhat similar environment are that the main router and a node is on 388 firmware, the backhaul is wifi, and the testing device is connected to the node either via Ethernet or wifi.

I find Tv streaming fine due to the amount of buffering but if you’re regularly streaming games from one device to another or doing something with minimal or no buffer and haven’t noticed it then yeah that probably confirms it’s either my environment or the gt6s specifically that are the issue.

Just coming back to this in case anyone else has the same issue as I think I have resolved the problem, although the solution is not great.

Thank you Saltgrass for your suggestions, they were very helpful. I tried turning every wificapable device off but sadly that didn’t solve the problem.

Quick summary of the issue:

  • Network consists of 4 GT6 routers, one main router and three aimesh nodes using 5Ghz dedicated backhaul
  • Consistent bad lag spikes on all devices lasting 3-10seconds occurring every five minutes / 300 seconds when connected to an aimesh node in any way.
  • No issue on devices connected directly to the main router
  • Lag spikes occur on every device on every node at exactly the same time. Power cycling a node does not change the timing. Power cycling the main router does change the timing
  • Changing which gt6 is the main router doesn’t fix the issue, nor does a hard reset and setting back up again without a backup settings file.
  • Could not find any setting to change on the wifi backhaul settings that fixed the problem - channel, bandwidth, professional settings, etc, tried making no modifications from default settings after hard reset
  • No log issues noted; no obvious processes running at the time of the issue when using ssh to access the main router; killing numerous processes running at the time of the issue did not solve the issue; trying to modify the right Wi-Fi conf inactivity polling for the hostapd service did not fix the issue 
  • Ethernet backhaul seems to eliminate the problem on that node but not a practical solution and defeats part of purpose of triband routers
  • Trying my old ax82u and ax3000 (running Merlin) as a seperate network they had no problem.
  • Downgrading main router only firmware from 24580 firmware to 24512 did not fix the problem

 

The solution was downgrading the main router firmware only further down to 23145.

So between 23145 and 24512 there must have been a code change that introduced this issue. I suspect any network using a gt6 as the main router, recent firmware and a wifi backhaul may experience the same issues with nodes but can’t confirm that.

 

As I say sadly not a great solution, particularly if the units decide they will selfupgrade the firmware occasionally and aren’t able to get security updates, so I’m still considering returning all four units, despite the speeds and stability I’m getting outside of this issue being great.

 

 

Hi Cloakedrat,

Where I found old firmwares for GT6? I have 2 nodes GT6 with latest firmware. Then I cannot save my changes. For example, change a 5 GHz-2 WiFi frequency 116 to 108, click save then show progress bar but finally i see old automatic settings.

Hey Jarvisit,

That is really strange, don’t think I’ve noticed that one myself. I found the old firmware here:

https://drivers.softpedia.com/dyn-search.php?search_term=Rog+rapture+gt6+firmware

I would have liked an official source but this is the best I could find.

i tried contacting ASUS support about the 5minute node lag spike issue so they can flag with the appropriate firmware team but they seem unable/unwilling to be able to deal with software issues, insisting the devices all be sent in to be repaired. This is despite trying and seeing a couple of times that the issue goes away when downgrading to 23145 and comes back when upgrading to 24512+

If anyone else with gt6s, latest firmware and wifi backhaul is able to see if they have the same issue with devices connected to a node that would be really helpful to confirm if im just unlucky with these specific devices or it’s a wider issue.