RB4011: wlan1 disabling itself

TimurA, my friend, thanks a lot. Seems like turning the local forwarding off fixed my problems with more than 20 CAP ACs.

always please skb73.

rechecked. Unfortunately wlan1 only works through CAPSMAN without Local forward.

I confirm.
With CAPsMAN work perfect.
3 days work perfect without problem.

Can you please post you config on a working RB4011 running with CAPSMAN?

I cannot explain it, but now it works for 3 days without capsman… Still waiting for a glitch



/interface bridge port
add bridge=bridge1 interface=ether2
add bridge=bridge1 interface=ether3
add bridge=bridge1 interface=ether4
add bridge=bridge1 interface=ether5
add bridge=bridge1 interface=ether6
add bridge=bridge1 interface=ether7
add bridge=bridge1 interface=ether8
add bridge=bridge1 interface=ether9
add bridge=bridge1 interface=ether10
add bridge=bridge1 interface=wlan2

/interface wireless cap
# 
set bridge=bridge1 discovery-interfaces=bridge1 enabled=yes interfaces=wlan1

/caps-man configuration
add country=greece datapath.bridge=bridge1 hide-ssid=yes installation=indoor \
    keepalive-frames=enabled mode=ap multicast-helper=default name=cfg1 \
    rates.vht-supported-mcs="" security.authentication-types=wpa2-psk \
    security.group-encryption=aes-ccm security.group-key-update=5m \
    security.passphrase=123456 ssid=MySSID

/caps-man interface
add configuration=cfg1 disabled=no l2mtu=1600 mac-address=12:34:56:78:90:12 \
    master-interface=none name=cap1 radio-mac=12:34:56:78:90:12 radio-name=\
    012345678912

/caps-man manager
set enabled=yes

/caps-man manager interface
set [ find default=yes ] forbid=yes
add disabled=no interface=bridge1

/caps-man provisioning
add action=create-dynamic-enabled disabled=yes master-configuration=cfg1

so, technically, me, who didn’t use all mk gimmicks, just plain setup… i had a perfectly useful router! :slight_smile:)

Well, knock on wood, but after last round of config changes my 5G wifi is still working with 10 days of uptime…

[admin@MikroTik] > /system resource print
             uptime: 1w3d12h10m4s
            version: 6.44.5 (long-term)
         build-time: Jul/04/2019 10:32:21
   factory-software: 6.43.10
        free-memory: 955.2MiB
       total-memory: 1024.0MiB
                cpu: ARMv7
          cpu-count: 4
      cpu-frequency: 1400MHz
           cpu-load: 0%
     free-hdd-space: 128.8MiB
    total-hdd-space: 170.9MiB
  architecture-name: arm
         board-name: RB4011iGS+5HacQ2HnD
           platform: MikroTik
[admin@MikroTik] > /interface wireless monitor [find] once
                   status: running-ap           running-ap
                  channel: 2447/20-Ce/gn(17dBm) 5200/20-eCee/ac/P(20dBm)+5530/80/DP(24dBm)
        wireless-protocol: 802.11               802.11
              noise-floor: -102dBm              -105dBm
           overall-tx-ccq: 77%                  100%
       registered-clients: 2                    1
    authenticated-clients: 2                    1
              wmm-enabled: yes                  yes

I can’t suggest too much aside of 6.44.5 but I currently have it set up in 80+80 mode with statically configured frequencies. Specifically

set wlan1 band=5ghz-n/ac channel-width=20/40/80mhz-XXXX frequency=5200 secondary-channel=5530

I do have

adaptive-noise-immunity=none

(and I remember ANI being a massive pain on OpenWRT in the past) but it was off the whole time before anyway.

I actually had a massive amount of trouble setting up 80+80.
For first bunch of attempts “set secondary-channel=” command was accepted, returned me to prompt… but crashed the wireless subsystem.
Both 2.4 and 5.0 wifi go down, and many commands such like /interface print (not to mention /interface wireless print) just frozen the console.
A minute or two later I’ve been also losing the ability to open new ssh connections - key was accepted but then terminal hanged in blank.

Only fix for that was a power cycle.
Actually, no, that was a long night and I forgot the whole story.
Because the faulty config was already saved, after a reset I could log in but wifi was still crashed and trying to do e.g. /interface print would freeze the management plane again.
So after that power cycle it was actually required to restore backup, which would then say it will restart but fail to do so (because crashed system components).
Manual power cycle at that point brought me back to the config from backup, so with the ability to edit wireless config again.
So, try another setting, freeze the device again, rinse, repeat.

I’m still not 100% sure but I think what resolved this was setting

channel-width=20/40/80mhz-XXXX

.
Earlier it was specifically

channel-width=20/40/80mhz-Ceee

on the assumption of “if I’m setting frequencies manually, I should set up the control channel position manually too, right?”.

But… yeah. Works for me for now, at least.

PS. Can anyone point out to me what’s the logic for the code-tag to sometimes render inline and sometimes as a box?

Nope. Playing with channel settings still have no effect.

I’ll stay with CAPsMAN until MT release a patch in stable ROS.
In this time, is the only solution that work 100%.

It was discovered empirically that macbook pro 15 “2017 and macbook pro 15” 2019 are to blame for the fall of wlan1. There is no such effect on windows 10.

this is in my case.

ps
maybe it’s itunes wifi synchronization with iphone, ipad behaves like this or Airdrop or something specific - which I don’t know.

ps2

disabled AirDrop everywhere until the interface crashes. I will watch.

After a year of work, my HAP AC2 failed. Bought to replace RB4011. And in this device, for 3 days of operation, WiFi turned off 2 times. Does everyone have a problem or replace the device in the store?

@aya: Please, respect each other. You posted in a topic where multiple people are describing the problem for 4 pages - you should take a look first I think.

The MT is working on the solution. Replacing the device will not help.

In addition… in this time we have a solution that work perfect without any problem…
The solution is CAPsMAN and i have a mini tutorial / sample code above.

Please write to support@mikrotik.com, providing supout.rif from your device generated after the crash (when the interface is “down”). Additionally, If you know how to repeat the issue with your setup, please provide detailed steps.

Several days passed, still no replies and not a single email about this issue.

Is the RB4011 5GHz issue resolved?

Hello Normis,

No, the problem is not resolved.
we sit and watch you :slight_smile: when something is decided :slight_smile:

PS
I sent new data to Reiniss, and my conclusions about the SSID drop.

Timur, I understand from Reinis, that your problem is not related to wireless at all, it’s most likely a Layer 2 misconfiguration problem. My question was to the other participants of this thread.

Normis!

here it is not a level 2 configuration. You have some problems with RoS on L2. Since it affects wlan1 but not wlan2 - how can you explain this? When all the interfaces are in one bridge! And checked without a bridge! Let's not throw off the problem for a special case!

You DanielJB and others described the same symptoms. ping increases and SSID translation stops

/caps-man remote-cap print

ADDRESS NAME STATE RADIOS

2 172.27.0.1/58405 [11:E9:4B:BF:AB:CE] Run 2

Explain to me what is going on on L2 in RB4011. When the device in capsman constantly generates a left mac address. Non-native!
I already mentioned this to Arthus and Rudolfs and Reinis.

PS
Remember how we resolved the issue of identical mac addresses on SFP and wlan1 for a long time.