After moving from provider router to the new Flint 3 about 2 months ago I noticed sudden connection drops. Usually it was sudden short connection breakage affecting e.g. YouTube streaming and restoring shortly afterwards.
I have looked in the logs after logging the time of breakages and noticed the same pattern: during this short connection I was having the same kernel errors as in this topic
Sat Nov 29 20:54:07 2025 kern.err kernel: [64421.638755] Sat Nov 29 20:54:07 2025 kern.err kernel: [64421.643027] wlan: [7312:I:ANY] wlan_cfg80211_change_station: Ignore set station for ap vlan wlan0 Sat Nov 29 20:57:57 2025 kern.err kernel: [64651.526690] wlan: [7312:I:ANY] wlan_cfg80211_change_station: Ignore set station for ap vlan wlan0 Sat Nov 29 20:58:24 2025 kern.err kernel: [64678.426187] wlan: [7312:I:ANY] wlan_cfg80211_change_station: Ignore set station for ap vlan wlan1 Sat Nov 29 20:58:26 2025 kern.err kernel: [64680.116042] wlan: [0:E:ANY] ieee80211_recv_asreq: ieee80211_recv_asreq: assoc req from dms not-supported sta : Sat Nov 29 20:58:26 2025 kern.err kernel: [64680.116042] Sat Nov 29 20:58:26 2025 kern.err kernel: [64680.120720] wlan: [7312:I:ANY] wlan_cfg80211_change_station: Ignore set station for ap vlan wlan0 Sat Nov 29 20:58:35 2025 kern.err kernel: [64689.759438] wlan: [7312:I:ANY] wlan_cfg80211_change_station: Ignore set station for ap vlan wlan1 Sat Nov 29 21:02:04 2025 kern.err kernel: [64898.493560] wlan: [0:E:ANY] ieee80211_recv_asreq: ieee80211_recv_asreq: assoc req from dms not-supported sta :
Devices in question are both Apple Watches: Series 9 and SE. The YouTube connection breakage coincides with those kernel errors (YouTube is from iPad Pro 2018 on the same network, connected to 2.4/5ghz)
MLO is disabled, I just have 3 networks with the same SSID on different frequencies. Have not seen any instability on 6Ghz yet.
Disabling hardware acceleration has not helped.
Firmware is v4.8.3, the issue was present on v4.8.2 as well.
Router is scheduled to automatically reboot at night
Have anyone else encountered similar issues? Are there workarounds better than “reset and try again”?
Hmm it has something to do with AP vlans and wireless connectivity.
Have you setup this with multi psk perhaps?
I would think it must be a exotic change which trigger this.
The issue with such things is that 9/10 these Qualcomm or other vendor SDK are not feature rich, or they need different control/command which the OpenWrt backend doesn't understand in luci/uci/netifd/cfg80211, its really the minimum you can get from OpenWrts functions which is basic wifi.
You have to figure which option causes this error.
In my experience, any OpenWrt-ish device with constrained RAM/no swap benefits from regular reboots (particularly if unattended).
Or uptime limits, watchdog scripts, scheduled reboots - something.
This usually applies to devices at the lower end of the GL line. Not sure about Flint 3 - I one for a week and have very little experience with it beyond that short test period.
No fancy setup at all. It is as easy as it gets. Besides changing passwords and the reboot schedule I have lucl stats with ping statistics. I have configured dynamic dns. Also I had WireGuard server enabled for some time. Everything else was a complete stock. Regular wlan with the password. On the stable branch
In the thread I linked there were also issues with IOT low power devices. Probably some specifics there, since I guess Apple Watch should be using some low power mode
Change the 2.4 GHz SSID to a separate, IoT-dedicated network.
SSH into the router and run the command to set the 2.4 GHz operating mode to “11ng.”
uci set wireless.wifi0.hwmode='11ng'
uci commit wireless
ubus call network reload
Connect your IoT devices, including the Apple Watches, to this network.
After making these changes, please observe whether the issue persists.
Some IoT devices may have compatibility issues with newer technologies like Wi-Fi 7, which can lead to connection problems.
We also recommend updating the device firmware to see whether that resolves the issue.