Intermittent loss of Wi-Fi since upgrading to 386.7

[geobernd]

I uploaded a second RT-AX86U test build, this one using the 5 GHz firmware from 386_49599. Please compare both.

RT-AX86U_386.7_1-g8fd8deefbb_49599.w

Performance near and far from router is slightly better for me with this one (~10-% improvement - so just a hair above variation of measurements).
Also the UNII-4 channels have disappeared again from the dropdown...

I'll keep it running in 'production' with WiFi TV Streaming etc. tonight and will report back if any issues....

 

[RMerlin]

Flashed the new build 386.7_1-g8fd8deefbb_49599. Didn't notice the 5 Ghz driver was dated July 2021, but now it looks like this...

Yes, I don't know if Asus sent me the old driver by mistake or if it was intentional. Taiwan is already on their weekend, so I won't get any clarification until next week. I figured that in the mean time I'd let people test both to see if they see any difference in terms of stability.

 

[octopus]

I just loaded 386.7_1-g8fd8deefbb and I don't get any UNII-4 channels.

 

[Calkulin]

If it works, yes. It was one of the reasons I haven't released 386.7_2 yet as I wanted to see if obtaining a fixed driver was possible without requiring a new GPL.

It's not guaranteed to fix every single wifi issue, as some may be related to configuration, not the driver itself. This is most specifically related to connection stability, not to channel selection behaviour, which is handled by a different component.

Do you know if they are also going to send you updated drivers for the RT-AX88U and GT-AX11000? Or are you just going to focus on fixing the RT-AX86U?

 

[RMerlin]

Do you know if they are also going to send you updated drivers for the RT-AX88U and GT-AX11000? Or are you just going to focus on fixing the RT-AX86U?

RT-AX88U yes, as indicated above because of the new SoC. Currently no plan to update the GT-AX11000 until the next major GPL merge.

I just loaded 386.7_1-g8fd8deefbb and I don't get any UNII-4 channels.

UNII-4 channels are only available in a few regions like the US.

 

[kernol]

I uploaded a second RT-AX86U test build, this one using the 5 GHz firmware from 386_49599. Please compare both.

RT-AX86U_386.7_1-g8fd8deefbb_49599.w

Solid on both my RT-AX86U's -

Top client device is latest (2022) iPad Pro 12.9" / Last client listed in Samsung S22+ android phone.

Will post if any issues encountered.

 

[DiGz_Au]

new test build for ax86u wifi is more stable now. thanks

 

[octopus]

UNII-4 channels are only available in a few regions like the US.

Thanks, didn't realise that.

 

[charlie2alpha]

The first test build for the AX86U was a bust, allowed my to select UNII-4 channels while it shouldn't (I'm located in Germany) and it wasn't doing the DFS check. The second build fixed all that, running stable so far.

 

[octopus]

I have compared this two wifi drivers, some findings.

386.7_1-g8fd8deefbb Displaying unii-4 cahnnels but no in drop down list. Seems have more transmitted power, and no DFS check.
386.7_1-g8fd8deefbb_49599 Doesn't display unii-4 cahnnels, trasmitted power is same as before, and have DFS check working.

When I go to Wireless =>Wireless MAC filter =>> then wifi crash and get BSSID: 00:00:00:00:00:00

Jul 16 09:20:30 wlceventd: wlceventd_proc_event(486): eth7: Disassoc 70:5F:A39:A0:C3, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8)
Jul 16 09:20:30 wlceventd: wlceventd_proc_event(486): eth7: Disassoc 3C:7C:3F:44:FD:24, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8)
Jul 16 09:20:30 wlceventd: wlceventd_proc_event(469): eth7: Deauth_ind 00:00:00:00:00:00, status: 0, reason: Deauthenticated because sending station is leaving (or has left) IBSS or ESS (3)
Jul 16 09:20:30 wlceventd: wlceventd_proc_event(486): wl1.2: Disassoc 88:3D:24:13:87:C0, status: 0, reason: Disassociated because sending station is leaving (or has left) BSS (8)
Jul 16 09:20:30 wlceventd: wlceventd_proc_event(469): wl1.2: Deauth_ind 00:00:00:00:00:00, status: 0, reason: Deauthenticated because sending station is leaving (or has left) IBSS or ESS (3)
Jul 16 09:20:30 kernel: wfd_unregisterdevice Successfully unregistered ifidx 1 wfd_idx 1
Jul 16 09:20:30 hostapd: eth7: STA 70:5f:a3:d9:a0:c3 IEEE 802.11: disassociated
Jul 16 09:20:30 hostapd: eth7: STA 3c:7c:3f:44:fd:24 IEEE 802.11: disassociated
Jul 16 09:20:30 hostapd: eth7: STA 00:00:00:00:00:00 IEEE 802.11: disassociated
Jul 16 09:20:30 hostapd: wl1.2: STA 88:3d:24:13:87:c0 IEEE 802.11: disassociated
Jul 16 09:20:30 kernel: br0: port 11(wds1.0.1) entered disabled state
Jul 16 09:20:30 kernel: device wds1.0.1 left promiscuous mode
Jul 16 09:20:30 kernel: br0: port 11(wds1.0.1) entered disabled state
Jul 16 09:20:30 hostapd: wl1.2: STA 00:00:00:00:00:00 IEEE 802.11: disassociated

 

[Calkulin]

RT-AX88U yes, as indicated above because of the new SoC. Currently no plan to update the GT-AX11000 until the next major GPL merge.

The RT-AX88U and GT-AX11000 also have the same wireless dropping issue that the RT-AX86U does, so I would of thought if you fixed one, you would fix all the models that have that issue but ok, thanks for clarifying

 

[iJorgen]

18 hours uptime on 386.7_1-g8fd8deefbb_49599 with no issues/drop-outs and clean logs, but I didn't have any issues before either.

 

[Threska49]

Dirty upgrades are always a roll of the dice. If you don't like having to do a factory refresh each time, it's good policy to NOT chase upgrades, esp. if done for no specific, known purpose other than remaining current. Frankly, I'm amazed at the number of users who will do a dirty upgrade when it isn't actually necessary, sometimes even remotely!

Unless I'm addressing a specific, known issue w/ the upgrade, I only update every 8-12 months. And it's always a clean, factory fresh install w/ manual configuration. I avoid all kinds of hassle that way. Don't let yourself be sucked in by claims of others on the release thread about how their dirty upgrade worked. Many times you won't even notice an issue until much later, then in frustration, you end up doing a complete reset anyway!

Well you're right about "roll of the dice" although my update worked well for a good while, then I noticed I had pages that didn't load on my phone. I had both a better ethernet connection, AND my phone did better connecting using a cellular connection. Kind of a mystery because otherwise all the clients will authenticate and connect to the host.

 

[Threska49]

Seems to me if the vast majority of those having issues w/ 386.7 are NOT having them w/ 386.5_2, then it makes no sense to continue w/ 386.7.

In my experience, far too many ppl upgrade just for the sake of remaining current. Unless the new firmware addresses serious vulnerabilities or bugs that directly affect you, you should STOP chasing upgrades just because they are available. You need to appreciate the value of stable existing firmware before delving into the unknown of new firmware, esp. if its only known benefit is that it's NEW!

As far as I can tell, there's simply no compelling reason to embrace 386.7 that warrants all this trouble. Esp. if 388.x supposedly offers some relief.

Well except from a security standpoint one doesn't know it will affect them till it happens. Pretty much all computers have this problem, even routers. Hence the exploits you hear on the news. In other words people upgrade for more than just features but security.

 

[dfunked]

Yep, I tend to automatically update stuff as soon as there's an update available. I'd rather hear about an exploit then and find out that I patched it weeks ago rather than rushing to get everything patched.

Of course there are times like this when you end up rolling back and applying a bit more caution before your next update, but they're rare enough for the cons not to outweigh the pros imo.

 

[cc666]

Solid on both my RT-AX86U's -
View attachment 42853
Top client device is latest (2022) iPad Pro 12.9" / Last client listed in Samsung S22+ android phone.

Will post if any issues encountered.

I cant find the screen that shows me the channel width by connection. I select Network Map - Then view list and I do NOT see the channel width connection?

CC

 

[cc666]

I cant find the screen that shows me the channel width by connection. I select Network Map - Then view list and I do NOT see the channel width connection?

CC

Disregard I found it.

CC

 

[QuikSilver]

RT-AX86U_386.7_1-g8fd8deefbb_49599.w seems to be solid on my AX86u

 

[TheLyppardMan]

I cant find the screen that shows me the channel width by connection. I select Network Map - Then view list and I do NOT see the channel width connection?

CC

Do you mean this?

 

[Damit1]

Tried to figure it out on my own- but did not manage.
AX88U is having WIFI dropping as well, any known beta for that matter?