"Disassociated because sending station is leaving (or has left)" - Roaming issue?

[CB7]

Hi,

I have another question. This one has been driving me crazy for quite some time now and has been a noticeable issue since deploying Merlin - but I think probably has been a problem before, but I just wasn't paying attention to it yet. I've found various threads on the matter both here as well as on Reddit (most recent I think is this one: https://www.snbforums.com/threads/a...ess-clients-dropping.88646/page-2#post-924146) and you have tons and tons of suggestions. "The problem is beamforming!", "The problem is your iPhone!", "You need to select channel 36!" and none of it actually helped. So I figured I could find out what device may be the culprit, but upon closer inspection I see it doing this over and over again with multiple devices:

Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:02 wlceventd: wlceventd_proc_event(662): eth6: Disassoc <MAC>, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8), rssi:0
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(685): eth6: Auth <MAC>, status: Successful (0), rssi:0
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(695): eth6: ReAssoc <MAC>, status: Successful (0), rssi:-58
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(685): eth6: Auth <MAC>, status: Successful (0), rssi:0
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(695): eth6: ReAssoc <MAC>, status: Successful (0), rssi:-58
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(685): eth6: Auth <MAC>, status: Successful (0), rssi:0
Nov 5 14:12:03 wlceventd: wlceventd_proc_event(685): eth6: Auth <MAC>, status: Successful (0), rssi:0
Nov 5 14:12:04 wlceventd: wlceventd_proc_event(695): eth6: ReAssoc <MAC>, status: Successful (0), rssi:-82
Nov 5 14:12:04 wlceventd: wlceventd_proc_event(695): eth6: ReAssoc <MAC>, status: Successful (0), rssi:-75

I found that rather interesting, as it disassociates with nearly half of the connected devices at once at exactly the same time; only a few of which are Apple. (In case it truly is the iPhone, but I don't have "private Mac" enabled to begin with, because I give static DHCP leases based on MAC-address - been doing so for years).

This begs the question: when it does the ReAssoc, it displays the RSSI. The -82 one for example certainly would be disconnected by it and -75 is on the border. (The -58 absolutely not).
... Is this problem simply caused by the Roaming Assistant? Or am I being too optimistic in hoping that nobody suffering from this issue in the countless of threads on it has thought of that?
I'd prefer to keep the Roaming Assistant on to give devices a little nudge out of the door when reception is on the border of the two AP's.


Thanks

 

[sfx2000]

client side power saving...

 

[Lodanum]

Hello

I had exactly the same problem with my ax11000. I was no longer able to connect my connected devices to the 2.4ghz band. He found the wifi and I always had the error message "the router does not respond". In the logs, I had the same error message as you. Attempts to connect and disconnect with the famous rssi. The solution I found to be able to connect my devices and therefore no longer have this error message in the logs was to rename the 2.4ghz band. The connection of the devices was directly operational. On the other hand, I still notice loss of wifi on all different frequencies occasionally and less good coverage. My router may be at the end of its life

 

[CB7]

Hello

I had exactly the same problem with my ax11000. I was no longer able to connect my connected devices to the 2.4ghz band. He found the wifi and I always had the error message "the router does not respond". In the logs, I had the same error message as you. Attempts to connect and disconnect with the famous rssi. The solution I found to be able to connect my devices and therefore no longer have this error message in the logs was to rename the 2.4ghz band. The connection of the devices was directly operational. On the other hand, I still notice loss of wifi on all different frequencies occasionally and less good coverage. My router may be at the end of its life

Thanks! Hmmmm.... My problems are primarily with 5Ghz, but indeed the 2.4Ghz band is using the same SSID. I switched away from seperate 2.4/5Ghz SSID's when I was making WiFi available in the garden, heh. Garden AP is connected through powerline adapter and doesn't do more than 50MBit, which is fine for streaming. But sometimes, the devices indoor would prefer the garden AP. So I disabled 5Ghz and only made it run on 2.4Ghz. This way: there's a seamless transition when going outdoors (switch to 2.4), but client indoors would always favor the 5Ghz band from the indoor routers and never even try to connect to the 2.4 signal. If I rename the 2.4Ghz network: I'd have connection drops when going from inside to outside, which isn't good for Teams and such. But for diagnostics purposes: I will give this a go to see if it indeed solves the problem and report back. (Probably somewhere next week.) Thanks!

client side power saving...

Thanks for your reply! I find that a bit hard to believe, they all disconnect at exactly the same time. Different devices, different vendors - but they do disconnect and reconnect in unison? But more importantly: it even disconnected my laptop in the middle of running a speedtest with that message. I... Don't think it will have gone in to powersavings while happily pushing out ~700MBit/sec (before it disconnected and errored out). You sure its all the connected clients doing this and not the router?