One of my Vero 4k devices keep losing the network connection. It doesn’t happen very fast - but after 10-15 minutes normally
I have tried to make a log of it. Video stopped at about 2021-02-10 23:12:xx.xxx
https://paste.osmc.tv/sufinesiro
I’m not really sure what to do. I do not have the option to permanently give this device a wired connection sadly.
I hope you can help.
Hi,
Only a guess, but I think yout current wifi channel maybe over populated. Suggest changing wifi channel on wifi router/AP. You can find out channel population using the suggestions here:
Thanks Tom.
WLC_E_DEAUTH_IND seems to suggest the router is forcing key renegotiation frequently.
Can you let us know what router you are using and the settings for your network?
Sam
Thanks, I have 3 Asus AC-86U setup with AiMesh2 - with ethernet backhaul.
There are no speciel speciel settings.
I have found the least populated wifi channels (there isn’t much going on here so)
I have another Vero 4k in another room - which doesn’t have this problem.
it’s also connected to the same node.
Hi,
Is the router running the latest firmware? If you are running stock, it looks like there was release a few days back.
Thanks Tom.
The log is showing a lot of “group rekeying messages” and the disconnect line shows:
Feb 10 23:12:46 osmcsove wpa_supplicant[243]: wlan0: CTRL-EVENT-DISCONNECTED bssid=d4:5d:64:db:03:b4 reason=16
Reason code is also mentioned further down:
Feb 10 23:12:46 osmcsove kernel: CFG80211-ERROR) wl_is_linkdown : Link down Reason : WLC_E_DEAUTH_IND
Feb 10 23:12:46 osmcsove kernel: link down if wlan0 may call cfg80211_disconnected. event : 6, reason=16 from d4:5d:64:db:03:b4
Feb 10 23:12:46 osmcsove kernel: CFG80211-ERROR) wl_is_linkdown : Link down Reason : WLC_E_DEAUTH
According to this, reason code 16 means “Group Key Handshake timeout”, so everything might be related. If you can increase the group rekey period on the router, things might improve.
Also one observation. When it starts up, connman checks if it has an Internet connection. The log shows that it took seven seconds for this check to succeed, so something somewhere wasn’t working well.
Feb 10 18:52:21 osmcsove connmand[332]: wlan0 {add} route 212.227.81.55 gw 192.168.1.1 scope 0 <UNIVERSE>
...
Feb 10 18:52:28 osmcsove connmand[332]: wlan0 {del} route 212.227.81.55 gw 192.168.1.1 scope 0 <UNIVERSE>
Just to be sure, I have another Vero 4K that’s connected to the same node - and that device doesn’t have these problems at all (also Wifi)
The “Group Key Rotation Interval” is set to 3600 (seconds) - I have not changed it to 86400.
There are 4 other devices (than the two Vero 4K) that’s also connected (phones and laptops) - they don’t have any problems with internet access or losing connection.
I have a feeling something is wrong with this Vero 4k 
I don’t have the beta version of the firmware installed - but the release from January and I installed it back then.
This problem just started this week.
I’d be more inclined to look at your mesh network.
1 While you might not have “much going on” at your location, you actually have three routers on the same channel (36).
2 When the Vero4K gets kicked off one access point (d4:5d:64:db:03:b4)
Feb 10 23:12:46 osmcsove wpa_supplicant[243]: wlan0: CTRL-EVENT-DISCONNECTED bssid=d4:5d:64:db:03:b4 reason=16
it then tries to (re)connect to a different access point (d4:5d:64:db:1b:7c)
Feb 10 23:12:46 osmcsove wpa_supplicant[243]: wlan0: Trying to associate with d4:5d:64:db:1b:7c (SSID='klovergig' freq=5180 MHz)
...
Feb 10 23:12:46 osmcsove wpa_supplicant[243]: wlan0: CTRL-EVENT-CONNECTED - Connection to d4:5d:64:db:1b:7c completed [id=0 id_str=]
3 Then very quickly back again to d4:5d:64:db:03:b4:
Feb 10 23:13:20 osmcsove wpa_supplicant[243]: wlan0: Trying to associate with d4:5d:64:db:03:b4 (SSID='klovergig' freq=5180 MHz)
4 Then this strange event that I don’t really understand but involves two APs:
Feb 10 23:18:30 osmcsove wpa_supplicant[243]: wlan0: WPA: Group rekeying completed with d4:5d:64:db:03:b4 [GTK=CCMP]
Feb 10 23:19:53 osmcsove kernel: CFG80211-ERROR) wl_is_linkdown : Link down Reason : WLC_E_DEAUTH_IND
Feb 10 23:19:53 osmcsove kernel: link down if wlan0 may call cfg80211_disconnected. event : 6, reason=16 from d4:5d:64:db:1b:7c
Feb 10 23:19:53 osmcsove kernel: CFG80211-ERROR) wl_notify_connect_status : BSSID of event is not the connected BSSID(ignore it) cur: d4:5d:64:db:03:b4 event: d4:5d:64:db:1b:7c
Feb 10 23:20:04 osmcsove wpa_supplicant[243]: wlan0: CTRL-EVENT-DISCONNECTED bssid=d4:5d:64:db:03:b4 reason=16
All in all, these problems seem to be related to your mesh system. Perhaps the problem Vero4K lies close to the radio boundary between the two devices.
Thanks for the feedback.
Not to be arrogant or anything - but I really don’t know what to do.
Regarding number 1: Mesh is using the same channels on all nodes - I cannot change that. As far as I understand that’s also how it’s supposed to work.
I’ve had this setup for months (obviously upgraded firmware etc on devices and router)
Hi,
Could test @dillthedog’s theory and swap the vero4ks over. Obviously not a final solution, but hopefully confirm what we are dealing with.
Thanks Tom.
Yes, that’s how a mesh system works but it’s a bit like having three neighbours competing for the same channel.
@Tom_Doyle’s suggestion about swapping over the V4Ks is a good one. Right now, there seems to be too much bouncing between two access points. WiFi problems such as this can sometimes be reduced by changing the position of the V4K or, possibly, the position of one of the access points.
This would be a good idea
I’d also suggest beforehand verifying that both devices are on the same version.
Sam
They are on the same version.
“Group Key Rotation Interval” is set to 86400 (seconds) now (24 hours) and I have “bound” the device to the nearest Aimesh router so it will not change to another.
After that I cannot get the problem anymore.
Thanks for your help!