Deauthenticated because sending station is leaving - AC68U

[Draken]

Trying to get some stability into my internal network and today once again the repeater (AC68U) disconnected from my main router (AC86U).

Both are running the latest 384.16 Alpha release as I had the same stability issues with the latest stable.

Multi-User MIMO, Beamforming have been disabled in case they were to blame but the issue doesn't seem to be those.

The error I am receiving just before disconnect is:

Deauthenticated because sending station is leaving (or has left) IBSS or ESS (3)
and then a couple of
Class 3 frame received from nonassociated station (7)

For around 15-20 secs when it decides to reconnect.

Hardware failure is highly unlikely as the AC68U was running for more than 5 years without any issues and the AC86U is brand new.

Any ideas appreciated cause it is really annoying. AiMesh is out of the question as well as it is very unstable.

And yes I do live in an area with loads of wifis.

 

[Butterfly Bones]

Trying to get some stability into my internal network and today once again the repeater (AC68U) disconnected from my main router (AC86U).

Both are running the latest 384.16 Alpha release as I had the same stability issues with the latest stable.

Multi-User MIMO, Beamforming have been disabled in case they were to blame but the issue doesn't seem to be those.

The error I am receiving just before disconnect is:

Deauthenticated because sending station is leaving (or has left) IBSS or ESS (3)
and then a couple of
Class 3 frame received from nonassociated station (7)

For around 15-20 secs when it decides to reconnect.

Hardware failure is highly unlikely as the AC68U was running for more than 5 years without any issues and the AC86U is brand new.

Any ideas appreciated cause it is really annoying. AiMesh is out of the question as well as it is very unstable.

And yes I do live in an area with loads of wifis.

Known issue, many posts about this. Closed source, only ASUS can fix the coding mistake, RMerlin can do nothing. See second half of this response.
https://www.snbforums.com/threads/r...13-is-now-available.57860/page-35#post-515660

 

[Draken]

Known issue, many posts about this. Closed source, only ASUS can fix the coding mistake, RMerlin can do nothing. See second half of this response.
https://www.snbforums.com/threads/r...13-is-now-available.57860/page-35#post-515660

Thanks. I am reading my questions since I moved to Asus Merlin and I sound like a proper moaner but to my eyes this is a huge no no. I am getting this issue every few hours and especially while gaming it is properly annoying.

How can people recommend Asus Merlin or ASUSWRT when something as basic as this is not fixed. I would expect my router to have an uptime of months and not to restart every X hours.

Also I was under the impression that the ASUS software is completely open source. Wasnt aware that it was open source but terms and conditions applied. Not happy to say the least. Already wasted 4-5h of my life troubleshooting. It shouldnt be like that.

A few of the disconnections today (I had more actually but couldnt be asked copy pasting).


Mar 2 18:42:06 wlceventd: WLCEVENTD wlceventd_proc_event(401)
Mar 2 18:44:43 wlceventd: WLCEVENTD wlceventd_proc_event(449)
Mar 2 19:48:47 wlceventd: WLCEVENTD wlceventd_proc_event(449)
Mar 2 20:36:28 wlceventd: WLCEVENTD wlceventd_proc_event(420)
Mar 2 20:48:46 wlceventd: WLCEVENTD wlceventd_proc_event(386)

 

[Butterfly Bones]

Thanks. I am reading my questions since I moved to Asus Merlin and I sound like a proper moaner but to my eyes this is a huge no no. I am getting this issue every few hours and especially while gaming it is properly annoying.

How can people recommend Asus Merlin or ASUSWRT when something as basic as this is not fixed. I would expect my router to have an uptime of months and not to restart every X hours.

Also I was under the impression that the ASUS software is completely open source. Wasnt aware that it was open source but terms and conditions applied. Not happy to say the least. Already wasted 4-5h of my life troubleshooting. It shouldnt be like that.

Sorry you feel that way, most of here highly appreciate what RMerlin adds to Asuswrt code using the parts that are open source. Many Asuswrt coding mistakes are cleaned up by RMerlin where he can.

I clear these annoyances using syslog-ng to scrape my webgui logs clean.
https://www.snbforums.com/threads/scribe-syslog-ng-and-logrotate-installer.55853/

These scripts add so much to Asuswrt-Merlin, because RMerlin adds the incredibly powerful scripting functionality. These scripts make the Asus routers so much more powerful and useful.

Again
"Closed source, only ASUS can fix the coding mistake, RMerlin can do nothing."

More info on closed source in Asuswrt GPL releases to RMerlin.
https://www.snbforums.com/search/1311211/?q=closed+source&o=date&c[node]=42&c[user][0]=10954

There are other choices for Asuswrt replacement firmware, Tomato, there are more that a Google search will show.

 

[Draken]

Not you fault obviously nor anyones who has contributed to the project. You are to be respected for what you do although I would question my commitment to assist Asus if I saw bugs like that staying open for that long.

I am just a bit annoyed by all the 3rd party reviewers who would recommend ASUSWRT over other firmware when it has such a bug unfixed for at least 6 months. I attached the logs above. It makes it impossible for me to use the router as I need stability. I was going 90+ days with no issues on DDWRT. From a 250+ Euro hardware you expect it to do so.

Also it is partially my mistake as I should have researched better. I dont want and I dont have the time to spend another 5-6h min setting up Tomato on both routers (which I havent used in the past) nor can I go back to DDWRT as the AC86U is not supported.

I guess another return to Amazon is incoming and it is a pity cause it is a really nice router hardware-wise.

And out of curiosity are there many similar "hidden" surprises in the ASUS firmware that you are aware of?

 

[Deleted member 22229]

Those log entry's should not be affecting your stability. They are always there when a client comes and goes from the network. The bug is as far as i understand is just showing them in the log when before it was not written to the log. You can make it stop with SSH running these commands.


1. nvram set log_level=5
2. nvram commit
3. reboot

 

[L&LD]

@Draken as @Kal-EL points out, those log entries should not be affecting your stability.

The possibility of a hardware issue isn't ruled out by either by the 'in use for 5 years' or 'brand new' states of the routers. In fact, I would counter the opposite, just on 'probability' alone.

What steps have you taken to put both router and therefore your entire network in a good/known state? Are you simply flashing different firmware without fully resetting the routers (please see the M&M Config in the link in my signature below)?

Are you doing a full reset and then using saved backup config files afterward?

Have you at least tried an AiMesh setup following @OzarkEdge's excellent setup notes here? I have only set up one myself, but it is running without issue for weeks now using the link indicated.

The repeater mode is not the best way to use a router in a heavily used WiFi environment. This in itself may be causing issues for you. Neither is simply turning off (or on) settings 'just in case' they cause issues.

I suggest to get both your routers to a good/known state (the M&M Config and the Nuclear Reset guides in the link in my signature may help) and test for each change you do before doing wholesale changes and possibly coming to rushed conclusions.

Let us know a little more of how this network is configured (physically and settings-wise) and there may be more pertinent help offered.

If following a full/optimum M&M Config leaves the network in the same state, it may be a hardware issue for sure.

There are many options here for you to try and get what you have working. Returning the hardware may be warranted, but it may also leave you in the same position with a new router too.

 

[Draken]

Those log entry's should not be affecting your stability. They are always there when a client comes and goes from the network. The bug is as far as i understand is just showing them in the log when before it was not written to the log. You can make it stop with SSH running these commands.


1. nvram set log_level=5
2. nvram commit
3. reboot

Thanks Kal-EL but can you clarify sth please? This client is my second access point where half of my devices are connected. So at random moments you lose the internet and the network restarts on all attached devices which is obviously a problem (clarification only two of the above 5 entries was the MAC address of the router indeed).

@L&LD: The hardware failure cannot be ruled out I agree but I find it very improbable that after 5+ years of seamless running the second router just gave up once flashed with Asus Merlin.

In regards to updates I do a WPS NVRAM erase and resetup everything (or sometimes I use an old backup). I went through your M&M Config guides and this is pretty much what I do. Also as I will explain further down I dont care about the 2.4Ghz speed so B/G protection, Preamble Type etc are not applicable to my I believe. The only think I haven't done is format the JFFS partition.

AiMesh I cannot use as I had a bad experience with an AC88U which has now been returned to Amazon. The issue was lower speeds than my previous setup and general instability.

Unfortunately wifi is the only way forward around here. Due to the quality of wires the powerlines are excluded as I have tried them in the past, so whichever wifi solution gives me the best speed I will stick with it.

The setup is simple. Main AC86U in the main sitting room and AC68U in one of the bedrooms. 5Ghz is reserved only for the routers to talk to each other as no other device connects to 5Ghz. Distance is around 10m. Every other device roams between the two AP in the house as they both transmit on the same 2.4Ghz SSID. I dont care about the speed of the 2.4Ghz connection but rather about the speed of the 5Ghz connection as two desktops are connected to the AC68U via ethernet. It is seriously overcrowded with wifi signals. I am getting at least 15 from the main sitting room. 5Ghz is a bit better.

5GHz setup
Wireless mode: Auto
20/40/80 Channel
56 control channel (it was the best one when setting it up)
Extension auto

And from the advanced setting I have disabled Airtimes Fairness and Universal Beamforming but that didnt help much.

When initially set up I left it pinging for a 3-4 hours both to the main router and google.com and there were no dropped packages but the day after it dropped twice during the day while surfing/streaming so it is sporadic. When I refer to drops I am talking about the desktops which are connected to the Repeater so basically the AC68U disconnects momentarily from the AC86U and produces the logs attached above.

The RSSI is around -58 dBm which is more than decent so signal quality and reception on the 5Ghz band at least should not be an issue.

 

[OzarkEdge]

My 2xRT-AC86U AiMesh had a very quiet log. Then the router died, so now I have one 86U with the log filling up with:

Mar  2 12:19:07 wlceventd: WLCEVENTD wlceventd_proc_event(386): eth5: Deauth_ind A4:DA:22:29:52:C3, status: 0, reason: Class 2 frame received from nonauthenticated station (6)
Mar  2 12:19:07 wlceventd: WLCEVENTD wlceventd_proc_event(386): eth5: Deauth_ind A4:DA:22:29:52:C3, status: 0, reason: Class 2 frame received from nonauthenticated station (6)
Mar  2 12:19:07 wlceventd: WLCEVENTD wlceventd_proc_event(386): eth5: Deauth_ind A4:DA:22:29:52:C3, status: 0, reason: Class 2 frame received from nonauthenticated station (6)
Mar  2 12:19:19 wlceventd: WLCEVENTD wlceventd_proc_event(420): eth5: Auth A4:DA:22:29:52:C3, status: 0, reason: d11 RC reserved (0)
Mar  2 12:19:19 wlceventd: WLCEVENTD wlceventd_proc_event(420): eth5: Auth A4:DA:22:29:52:C3, status: 0, reason: d11 RC reserved (0)
Mar  2 12:19:19 wlceventd: WLCEVENTD wlceventd_proc_event(420): eth5: Auth A4:DA:22:29:52:C3, status: 0, reason: d11 RC reserved (0)

The offending device is a tiny 2.4 GHz cam about 160' from the router in another brick building. It is marginally out-of-range and causing the excessive log entries.

I thought about adjusting the Roaming Assistant 2.4 GHz RSSI threshold, but have not bothered since the problem will go away once I re-deploy a node out that way.

OE

 

[Draken]

Mar  2 20:48:34 wlceventd: WLCEVENTD wlceventd_proc_event(386): eth6: Deauth_ind F0:79:XX, status: 0, reason: Deauthenticated because sending station is leaving (or has left) IBSS or ESS (3)

Unfortunately it is not the case for me.

 

[Deleted member 22229]

Thanks Kal-EL but can you clarify sth please? This client is my second access point where half of my devices are connected. So at random moments you lose the internet and the network restarts on all attached devices which is obviously a problem (clarification only two of the above 5 entries was the MAC address of the router indeed).

I have no doubt that client is unstable. But is it because the router is causing a disconnect, reconnect issue? Perhaps the client is dropping and reconnecting and the router log is just showing what it see's. no fault of it's own. Just saying that has to be known.

Those log entries are very annoying but i have never seen a client drop when using it. Thats why i disabled that event from spamming the log.

 

[Draken]

I have no doubt that client is unstable. But is it because the router is causing a disconnect, reconnect issue? Perhaps the client is dropping and reconnecting and the router log is just showing what it see's. no fault of it's own. Just saying that has to be known.

Those log entries are very annoying but i have never seen a client drop when using it. Thats why i disabled that event from spamming the log.

Obviously it shouldnt be happening at the first place but is there a way to track it down to the router or the client?

There is no event at 20:48 for the client so my educated guess is that the client is disconnecting rather than the router dropping the connection but then again I will wait for the experts.

 

[Deleted member 22229]

There is no event at 20:48 for the client so my educated guess is that the client is disconnecting rather than the router dropping the connection but then again I will wait for the experts.

Yes wait for the experts because that i am not. Just a long time Asus router user. Good luck !!

 

[Val D.]

56 control channel (it was the best one when setting it up)

What country are you in? In most countries 56 is a DFS channel.

 

[Draken]

What country are you in? In most countries 56 is a DFS channel.

Changed it to 112 but I dont think it will have much difference. Will monitor the disconnections over the weekend and let you know.

 

[L&LD]

112 is still a DFS channel (I think).

 

[Draken]

Changed it to 112 but I dont think it will have much difference. Will monitor the disconnections over the weekend and let you know.

Yeah you are right. Left it on autoselect excluding DFS so it is at 42 atm.

 

[Draken]

And its gone, repeater disconnected. After 30 mins of being online...

 

[ColinTaylor]

Try changing the repeater from 5GHz to 2.4GHz @ 20MHz bandwidth. Yes I know the link will be slower, it's just an experiment to see if it's stable.

 

[Draken]

Try changing the repeater from 5GHz to 2.4GHz @ 20MHz bandwidth. Yes I know the link will be slower, it's just an experiment to see if it's stable.

Thanks will try although wont help much as there is no point running it at 2.4GHz. Unfortunately no time the next 5-6 days. Just had a few mins to upgrade to the latest Alpha version. No change. Keeps disconnecting.