Can't get wifi interfaces on hAP ax S out of "bound" condition

Powered up an AX S I have which took a nose-dive some weeks ago from 3m high :man_facepalming:
Anyhow, turns out it still works (apart from the case being a bit loose on one corner, nothing else seems to be broken ?).

I applied your last config snippet and first thing I see, is waiting for DFS (1 minute).
After 1 minute, SSID testtest showed up in wifi analyzer on my laptop as CH64.
I could connect to it without any problem.

That was however using 7.23rc3 (it was a bit behind and I haven't touched it anymore since that drop).
Applied 7.24rc3, reboot and ... same.
It simply shows up and I can connect to it.

[...@hAP] > /interface/wifi/print detail
Flags: M - MASTER; B - BOUND; X - DISABLED, I - INACTIVE, R - RUNNING
0  MBX default-name="wifi1" name="wifi1" mac-address=xyz arp-timeout=auto radio-mac=xyz
 
1  MBR default-name="wifi2" name="wifi2" mac-address=abc arp-timeout=auto radio-mac=abc
       configuration.ssid="testtest" .mode=ap .country=Germany security.ft=yes .ft-over-ds=yes .authentication-types=wpa2-psk,wpa3-psk
       channel.frequency=5320 .band=5ghz-ax .width=20/40/80mhz .skip-dfs-channels=10min-cac

So I still am not sure you are "detecting" that network in a correct way ?

Other thing I just think about: if your AC Lite is not able to use wpa3, best to remove it from the AP settings, only leave wpa2.
Some devices are pretty picky about that ...

Don't you have other clients to test ? Laptop ? Smartphone ? Something with a bit of recent wifi driver ?

Right now, I don't have anything else to try to detect with. The device is off-site in "production" (except the wifi bit, obviously), but as I mentioned, the whole thing was set off because users started complaining that they first had wifi reconnection issue and then didn't see the wifi any more. So if the configuration worked on your device, it's certainly not a config issue, and I'm looking at either a detection issue (possible) or a hardware issue (also possible), and given the user complaints, I'm leaning towards the latter for the time being. I'll try to get some other device in for testing though.

Thanks a lot, much appreciated!

Cheers,
Toby.

FWIW also connected to it with a Samsung S25.
My ISP connection is 1000/100 (or about), I got 565/98 using Speedtest (with CPU on AX S inching real close to 100% on that download).

So config is definitely not the base problem.

Yeah, I agree. Starts to smell like a hardware issue to me. I've contacted the MT support and sent them a supout. Let's see what they think.

Thanks a lot & cheers,
Toby.

The only thing I can suggest here (not owning an "S") would be to compare the listed capabilities in the "Radios" tab between the two. I note that trying to set something that is not supported by the specific radio can cause silent fail like this (hit that when I got my first AX devices and doing a freq plan.)

Your config seems simple enough, thatI can't see anything obvious though.

Do things behave any differently if you just set bare minimums in the config?

(Seems like other things are pointing to HW by now . . . replied prior to full read . . . . Looking at capabilities in Radios may still have merit, though . . . It's about the only place I know to see radio capabilities.)

Not sure I understand correctly. Comparing the capabilities of what two devices? My S and holvoetn's S?

Yes, I agree that you can (and I'd say, rather easily ;-)) create a configuration that breaks hits some subtle limitation. That's why I deliberately tried to keep the test config as open and simple as possible so I wouldn't hit any subtle traps.

Not quite sure what you mean by "bare minimums in the config"? I don't really see what could be thrown out after a /interface/wifi/reset and the above minimal testing config?

But the capabilities thing is certainly worth a go. Here's what my S claims it can do:

[admin@JSS10-Router] /interface/wifi/radio> print detail
Flags: L - LOCAL
0  L radio-mac=D0:EA:11:2D:EA:6D tx-chains=0,1 rx-chains=0,1 bands=2ghz-g:20mhz,2ghz-n:20mhz,20/40mhz,2ghz-ax:20mhz,20/40mhz
     ciphers=tkip,ccmp,gcmp,ccmp-256,gcmp-256,cmac,gmac,cmac-256,gmac-256 min-antenna-gain=6 countries=all
     2g-channels=2412,2417,2422,2427,2432,2437,2442,2447,2452,2457,2462,2467,2472 max-vlans=4095 max-interfaces=16 max-station-interfaces=19 max-peers=120 hw-type="mt7916"
     hw-caps=sniffer,qos-classifier-dscp,channel-switch,6,meshpoint interface=wifi1

1  L radio-mac=D0:EA:11:2D:EA:6E tx-chains=0,1,2 rx-chains=0,1,2
     bands=5ghz-a:20mhz,5ghz-n:20mhz,20/40mhz,5ghz-ac:20mhz,20/40mhz,20/40/80mhz,20/40/80/160mhz,5ghz-ax:20mhz,20/40mhz,20/40/80mhz,20/40/80/160mhz
     ciphers=tkip,ccmp,gcmp,ccmp-256,gcmp-256,cmac,gmac,cmac-256,gmac-256 min-antenna-gain=6 countries=all
     5g-channels=4920,4940,4960,4980,5180,5200,5220,5240,5260,5280,5300,5320,5340,5360,5380,5400,5420,5440,5460,5480,5500,5520,5540,5560,5580,5600,5620,5640,5660,5680,5700,5720,     
        5745,5765,5785,5805,5825,5845,5865,5885 max-vlans=4095 max-interfaces=16 max-station-interfaces=19 max-peers=120 hw-type="mt7916"
     hw-caps=sniffer,qos-classifier-dscp,channel-switch,6,meshpoint interface=wifi2

THX everybody for the thoughts and discussion!

Cheers,
Toby.

I think he meant comparing the capabilities of AX S versus AC Lite (or whatever client).

As I already said, if you set WPA3 and the client device (in this case AC Lite) is not capable to use WPA3, it might not connect AT ALL.
So you need to strip down that AP connection to make sure the client CAN connect to it.

Or (also indicated) use another client which CAN handle AP's capabilities.

Start from the bare minimum, maybe even leave out security completely.
When it connects, build up again but taking into account the client capabilities.
But when reaching this first point, you would already know the wifi works.

Ah. I see. Well, I'm not even trying to connect, am I? If I was trying to connect, then yes, I would think WPA2 vs. WPA3 would make a difference, but so far, I cannot even see the SSID? Wouldn't that be expected to show up regardless of whether it was using WPA2 or WPA3 in a scan?

And, of course, I'll try a different "client" as soon as I can, but unfortunately, the S is remote, so the immediate options are limited.

Cheers,
Toby.

You have a point there ... it should at least show.

And the other way around ?
Can you starting from AX S see that AC Lite ? Or any other AP broadcasting ?
Try both 2.4 and 5GHz channels.

As it happens, I did look at the AC's capabilities as well:

[admin@JSS10-AP] /interface/wireless/info> hw-info wlan2
      ranges: 4920-6100/5/a,an20,an40,ac20,ac40,ac80
   tx-chains: 0
   rx-chains: 0
  extra-info: pciinfo:0x0, cid:0, gain:2

So I do think that I did try mutually-possible channels. /shrug

As for "the other way around": I did try scanning with the S as well. Doesn't see anything at all. Nada. I mistakenly didn't mention that before as I was working on the hypothesis that somehow the S's radio was blocked by some weird config because of the bound flag, which turned out to be a wrong hypothesis.

So, yeah, the S doesn't see anything (on either interface).

Cheers,
Toby.

Hmm ... that's definitely not ok then.

Except for maybe other suggestions coming in, it might be worthwhile to create supout.rif on that AX S and shoot in a support ticket. Maybe support staff can see something in there which we do not.
Make sure to mention you tested both sides and that on both sides, nothing is detected.

Yeah. I was kinda hoping that there was some obscure config kink that was blocking the interfaces for some reason or another, but... So far, MT support has asked me to upgrade to -devel, which I did and which did not help, and asked for the supout file. Waiting to hear back from them... Apparently it's not a trivial problem or they would have come back to me already, I guess... XD

Anyway, THX again!

Cheers,
Toby.