Release Asuswrt-Merlin 386.5 is now available

[Wistuplu]

Thank you RMerlin, Asus, all the testers and addon developers for this wonderful platform.

386.5 is running fine since a few hours, after a dirty upgrade from 386.4 on AC86U.

The upgrade didn't work the first time, even though I had stopped the automated execution of connmon and spdMerlin addons; unmounted the USB stick and physically disconnected the USB stick. After like 30 minutes, I dared hitting the physical power button, rebooted and then the upgrade went fine. Well, as the README says:

"If the firmware upgrade fails, try rebooting your router to free up sufficient memory, without any USB disk plugged in, then try flashing it again."

Next time, I will first reboot before the upgrade, to avoid freaking out for 30 minutes while the router was kind of hanging with most LEDs ON, and the whole family enquiring about wifi.

Take care and thank you again.
Best regards

 

[palatouche]

I did some tests with these settings. Changed Default message log level from notice to warning and Log only messages more urgent than from debug to notice.
Rebooted and no luck, nothing changed. Always getting "not mesh client, can't delete it".

I saw a reply from RMerlin on another post and he said this:
"Default message log level is the level used by the router's own logging. Unfortunately, Asuswrt uses the same level for all of the log entries it generates itself, rather than distinguishing between notices, and critical errors when logging. This can be problematic for people using a centralized syslog daemon, as by default everything carried a priority of emergency, making it impossible to filter things. This setting allows you to assign a more sane default level so not everything from the firmware comes in as being an emergency.
The various daemons and services (dnsmasq, miniupnpd, etc...) will typically use an appropriate level when generating log entries.

Log only messages more urgent than is the default filter level for the router's own syslog (what is shown on the System Log page). If set to "All", then everything sent to the router's syslog gets shown. If set to a different level, then you can ensure that, for example, only errors get shown (once again, see the limitation in the previous paragraph regarding the firmware's own log entries)."

I did one last test: set Log only messages more urgent than to warning and the messages are gone.

 

[anotherengineer]

Maybe we should be talking about these settings?

View attachment 40060

Is there any relief to be found by tuning these? Mostly, I mean like you're saying above... many users don't even understand what they're reading in the logs, but they see keywords like "failed", "deauth", "tainted" and they lose their sh*t. Is there a log setting for "...just want to know if it's affecting the performance of my WAN or LAN?"

Is there any consensus then on a more "preferred" settings??

edit - set to this for trial & error

 

[palatouche]

Is there any consensus then on a more "preferred" settings??

edit - set to this for trial & error

Problem is we don't know what these settings really do ? Anyway give feedback on your tries.

 

[dev_null]

I'm curious about these entries in the log as well. I don't have RADIUS enabled as far as I can determine looking at wireless settings.

Mar  8 20:42:05 hostapd: eth6: STA 18:4e:cb:88:dd:fd IEEE 802.11: associated
Mar  8 20:42:05 wlceventd: wlceventd_proc_event(534): eth6: Assoc 18:4E:CB:88:DD:FD, status: Successful (0)
Mar  8 20:42:05 hostapd: eth6: STA 18:4e:cb:88:dd:fd RADIUS: starting accounting session 0A2B1A38056E216C
Mar  8 20:42:05 hostapd: eth6: STA 18:4e:cb:88:dd:fd WPA: pairwise key handshake completed (RSN)

Guys, i see a lof ot those below...

what are those messages?

Mar 9 17:41:07 kernel: XXXXXXXXX0 not mesh client, can't delete it

didnt see them before when i was on 3.2

But i just want to know what's they means. And if they are meaningless it's just garbage filling out the log and masking meaningfull entries.

If y'all are using Scribe (or are willing to do so), I described my method of scrubbing the log of extraneous and unwanted entries here: https://www.snbforums.com/goto/post?id=747000

 

[Dezzo]

You would know best, it's your network. You said here that you've always had issues on 5GHz-
View attachment 40093

Then you also had the issue, at least in January, on 386.3_2-
View attachment 40094

I'm not a wifi expert, but I do know faulty devices on cheap consumer ethernet switches like what's in your (and my) router do not hold up to faulty network adapters being on the network.

My recommendation to help your situation is to use whatever firmware version you prefer- but add 1 wireless device at a time, and add 1 more device every 2 days. See if one of your devices is causing an issue. Given you've always had problems, this seems like the methodical and prudent path forward.

I've recently had issues, and my issue was resolved by using Asus Firmware Restore to do an install of 386.2_3 on my AC66U_B1, which cleared up a bunch of bizarre issues that factory resets did not resolve. Do both of these troubleshooting steps, and you have a very good chance of resolving your network issues.

If you are still not believing me, try another firmware altogether, like the official Asuswrt or DDWRT. If no issues there, then life is good. There's really nothing else to do here, too many people have massive amounts of nodes connected to routers running 386.5 without problems. So there's 3 things to try, report back. But it doesn't sound like your issue has anything to do with Asuswert-Merlin.

Yes I've had minor 5ghz issues on several ASUS routers over the years (as have others), but usually it only entailed the 5ghz network dropping just once or twice a month. But since the 386.4 and 386.5 releases it's been happening a couple of times a day, so it's quite different.

The Google Mini speakers were confusing the issue, since they tend to not work well when channel bonding is used.
Once I disabled channel bonding and moved them to the 2.4ghz at 20mhz, 386.3_2 ended up having great stability for a couple of weeks. Soon as I upgraded to 386.5 again, the 5ghz issues returned for the devices connected to it.

When a problem appears with a specific firmware, and then goes away when that firmware is removed, then returns a second time with that firmware, it would tend to indicate an issue with the firmware. But given how many variables are involved in networking, it's always hard to nail down anything for certain.

 

[alecmascot]

It only connects on http
I have removed scMerlin so now there are no addons

So it has been good for 24 hours so it looks like scMerlin is the culprit.
I will leave it another day to be sure.

 

[Neville]

Hi!

Dirty upgrade from 386.4 on both AC86U and AX56u. Previously had 5Ghz drops, seems resolved

 

[Wadadli]

Is there any consensus then on a more "preferred" settings??

edit - set to this for trial & error

Your post seemed to go over everyone's head. Good thing it was just a warning shot !

 

[GHammer]

If y'all are using Scribe (or are willing to do so), I described my method of scrubbing the log of extraneous and unwanted entries here: https://www.snbforums.com/goto/post?id=747000

Not wanting to ignore the entries, just want to know why if I do not have RADIUS enabled nor in use I have the RADIUS entries.

 

[mikezilla2]

updated ....

is there Anything to learn from these 16 pages So far ?

 

[skeal]

updated ....

is there Anything to learn from these 16 pages So far ?

Yep! Read the pages that pertain to your model.

 

[RMerlin]

When a problem appears with a specific firmware, and then goes away when that firmware is removed, then returns a second time with that firmware, it would tend to indicate an issue with the firmware.

It could also indicate a configuration issue. If for example you had a certain Wifi feature enabled but that feature didn't work previously so it made no difference, and that fature gets fixed by a latter release, and that feature happens to be incompatible with some of your clients. Not saying this is your case, just pointing out that it can be one.

 

[mikezilla2]

Yep! Read the pages that pertain to your model.

i Have an AX88U just like U , i guess it wasn't obvious but just wondered who had been following along and who had not - Not me.

 

[Daveo]

Upgraded from 386.3_2 to 386.5 on RT-AX88U.

Seeing this in logs...

Mar 10 20:43:06 RT-AX88U-8B40 kernel: potentially unexpected fatal signal 11.
Mar 10 20:43:06 RT-AX88U-8B40 kernel: CPU: 1 PID: 15732 Comm: dcd Tainted: P           O    4.1.51 #2
Mar 10 20:43:06 RT-AX88U-8B40 kernel: Hardware name: Broadcom-v8A (DT)
Mar 10 20:43:06 RT-AX88U-8B40 kernel: task: ffffffc02f2d54c0 ti: ffffffc01f198000 task.ti: ffffffc01f198000
Mar 10 20:43:06 RT-AX88U-8B40 kernel: PC is at 0xf721339c
Mar 10 20:43:06 RT-AX88U-8B40 kernel: LR is at 0x1dce0
Mar 10 20:43:06 RT-AX88U-8B40 kernel: pc : [<00000000f721339c>] lr : [<000000000001dce0>] pstate: 600f0010
Mar 10 20:43:06 RT-AX88U-8B40 kernel: sp : 00000000ffa095d8
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x12: 00000000000a2050
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x11: 00000000f64ff024 x10: 00000000000a23c4
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x9 : 00000000f64ffa1c x8 : 00000000000a287c
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x7 : 00000000f64ffa54 x6 : 00000000000a2876
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x5 : 0000000000000000 x4 : 00000000f64ffa00
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x3 : 0000000000000000 x2 : 00000000ffa095b4
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x1 : 000000000007d72a x0 : 0000000000000000
Mar 10 20:53:17 RT-AX88U-8B40 kernel: potentially unexpected fatal signal 11.
Mar 10 20:53:17 RT-AX88U-8B40 kernel: CPU: 1 PID: 28984 Comm: dcd Tainted: P           O    4.1.51 #2
Mar 10 20:53:17 RT-AX88U-8B40 kernel: Hardware name: Broadcom-v8A (DT)
Mar 10 20:53:17 RT-AX88U-8B40 kernel: task: ffffffc01fc24b00 ti: ffffffc029cc8000 task.ti: ffffffc029cc8000
Mar 10 20:53:17 RT-AX88U-8B40 kernel: PC is at 0xf6fb239c
Mar 10 20:53:17 RT-AX88U-8B40 kernel: LR is at 0x1dce0
Mar 10 20:53:17 RT-AX88U-8B40 kernel: pc : [<00000000f6fb239c>] lr : [<000000000001dce0>] pstate: 600f0010
Mar 10 20:53:18 RT-AX88U-8B40 kernel: sp : 00000000fff966a8
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x12: 00000000000a2050
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x11: 00000000f62ff024 x10: 00000000000a23c4
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x9 : 00000000f62ffa1c x8 : 00000000000a287c
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x7 : 00000000f62ffa54 x6 : 00000000000a2876
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x5 : 0000000000000000 x4 : 00000000f62ffa00
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x3 : 0000000000000000 x2 : 00000000fff96684
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x1 : 000000000007d72a x0 : 0000000000000000
Mar 10 20:53:19 RT-AX88U-8B40 kernel: httpds (16816): drop_caches: 1

I have AIProtection fully on and using pixelserv-tls (to secure the router admin panel).

Tried just disabling AIProtection, but still get some kernel messages.

Not disabled/uninstalled pixelserv-tls, as I think this would remove the SSL protection on the router's HTTPS connection.

Any ideas?

 

[Treadler]

Upgraded from 386.3_2 to 386.5 on RT-AX88U.

Seeing this in logs...

Mar 10 20:43:06 RT-AX88U-8B40 kernel: potentially unexpected fatal signal 11.
Mar 10 20:43:06 RT-AX88U-8B40 kernel: CPU: 1 PID: 15732 Comm: dcd Tainted: P           O    4.1.51 #2
Mar 10 20:43:06 RT-AX88U-8B40 kernel: Hardware name: Broadcom-v8A (DT)
Mar 10 20:43:06 RT-AX88U-8B40 kernel: task: ffffffc02f2d54c0 ti: ffffffc01f198000 task.ti: ffffffc01f198000
Mar 10 20:43:06 RT-AX88U-8B40 kernel: PC is at 0xf721339c
Mar 10 20:43:06 RT-AX88U-8B40 kernel: LR is at 0x1dce0
Mar 10 20:43:06 RT-AX88U-8B40 kernel: pc : [<00000000f721339c>] lr : [<000000000001dce0>] pstate: 600f0010
Mar 10 20:43:06 RT-AX88U-8B40 kernel: sp : 00000000ffa095d8
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x12: 00000000000a2050
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x11: 00000000f64ff024 x10: 00000000000a23c4
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x9 : 00000000f64ffa1c x8 : 00000000000a287c
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x7 : 00000000f64ffa54 x6 : 00000000000a2876
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x5 : 0000000000000000 x4 : 00000000f64ffa00
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x3 : 0000000000000000 x2 : 00000000ffa095b4
Mar 10 20:43:06 RT-AX88U-8B40 kernel: x1 : 000000000007d72a x0 : 0000000000000000
Mar 10 20:53:17 RT-AX88U-8B40 kernel: potentially unexpected fatal signal 11.
Mar 10 20:53:17 RT-AX88U-8B40 kernel: CPU: 1 PID: 28984 Comm: dcd Tainted: P           O    4.1.51 #2
Mar 10 20:53:17 RT-AX88U-8B40 kernel: Hardware name: Broadcom-v8A (DT)
Mar 10 20:53:17 RT-AX88U-8B40 kernel: task: ffffffc01fc24b00 ti: ffffffc029cc8000 task.ti: ffffffc029cc8000
Mar 10 20:53:17 RT-AX88U-8B40 kernel: PC is at 0xf6fb239c
Mar 10 20:53:17 RT-AX88U-8B40 kernel: LR is at 0x1dce0
Mar 10 20:53:17 RT-AX88U-8B40 kernel: pc : [<00000000f6fb239c>] lr : [<000000000001dce0>] pstate: 600f0010
Mar 10 20:53:18 RT-AX88U-8B40 kernel: sp : 00000000fff966a8
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x12: 00000000000a2050
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x11: 00000000f62ff024 x10: 00000000000a23c4
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x9 : 00000000f62ffa1c x8 : 00000000000a287c
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x7 : 00000000f62ffa54 x6 : 00000000000a2876
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x5 : 0000000000000000 x4 : 00000000f62ffa00
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x3 : 0000000000000000 x2 : 00000000fff96684
Mar 10 20:53:18 RT-AX88U-8B40 kernel: x1 : 000000000007d72a x0 : 0000000000000000
Mar 10 20:53:19 RT-AX88U-8B40 kernel: httpds (16816): drop_caches: 1

I have AIProtection fully on and using pixelserv-tls (to secure the router admin panel).

Tried just disabling AIProtection, but still get some kernel messages.

Not disabled/uninstalled pixelserv-tls, as I think this would remove the SSL protection on the router's HTTPS connection.

Any ideas?

Disable Pixelserv.
Enable Aiprotect, your job will be done.

The alternate strategy: just ignore the log……be happy!

 

[capncybo]

Does anyone have a Link or List that shows or clarifies EXACTLY what features/options actually turns OFF HW acceleration (Runner: Enabled - Flow Cache: Enabled)?
I'm asking because on my RT-AX68U the Flow Cache seems to disable itself when I run the built in Speed Test...
Actually it might just happen when I click QoS - WAN/LAN Bandwidth Monitor & after the Gauges start to move.
Perhaps it's all tied to the same BWDPI/Trend-Micro-ish Engine... which would be OK momentarily, if the"Bandwidth Monitor" isn't compatible...
But shouldn't Flow Cache re-Enable itself after Leaving that page???
Not sure if it's a glitch or simply a limitation.
But if we can have (Runner: Enabled - Flow Cache: Enabled) should we try to maintain HW acceleration?

Note: I also have the following add-ons installed.
Diversion v4.2.2 (Without Pixel Server)
Skynet v7.2.8
FlexQoS v1.3.0
WireGuard Mgr v4.14

But I don't think any of these would be turning off my Flow Cache, or could they???

 

[Martineau]

Does anyone have a Link or List that shows or clarifies EXACTLY what features/options actually turns OFF HW acceleration (Runner: Enabled - Flow Cache: Enabled)?
Note: I also have the following add-ons installed.
WireGuard Mgr v4.14

But I don't think any of these would be turning off my Flow Cache, or could they???

In the spirit of full disclosure, wireguard_manager does auto-disable Flow Cache for certain models i.e. RT-AX86U

E:Option ==> ?

    Router RT-AX58U Firmware (v386.5_beta1)

    [✔] Entware Architecture arch=arm


    v4.16b3 WireGuard Session Manager (Change Log: https://github.com/MartineauUK/wireguard/commits/dev/wg_manager.sh)
    MD5=9b95efac623bd16b2063d95a4a047a66 /jffs/addons/wireguard/wg_manager.sh

    [✔] WireGuard Kernel module/User Space Tools included in Firmware (1.0.20210124)

<snip>

    [✔] Flow Cache is ENABLED

<snip>

but you can manually change it using the fc [disable | enable|?] command

E:Option ==> fc disable

    Broadcom Packet Flow Cache learning via BLOG disabled.
    Broadcom Packet Flow Cache flushing the flows

    Flow Cache Disabled
    (Use 'vx' command to uncomment config option 'DISABLE_FLOW_CACHE' to DISABLE permanently)

 

[capncybo]

In the spirit of full disclosure, wireguard_manager does auto-disable Flow Cache for certain models i.e. RT-AX86U

E:Option ==> ?

    Router RT-AX58U Firmware (v386.5_beta1)

    [✔] Entware Architecture arch=arm


    v4.16b3 WireGuard Session Manager (Change Log: https://github.com/MartineauUK/wireguard/commits/dev/wg_manager.sh)
    MD5=9b95efac623bd16b2063d95a4a047a66 /jffs/addons/wireguard/wg_manager.sh

    [✔] WireGuard Kernel module/User Space Tools included in Firmware (1.0.20210124)

<snip>

    [✔] Flow Cache is ENABLED

<snip>

but you can manually change it using the fc [disable | enable] command

E:Option ==> fc disable


Broadcom Packet Flow Cache learning via BLOG disabled.
Broadcom Packet Flow Cache flushing the flows
    Disabled

Thanks, yes I had found that setting (was curious if that was a set it & forget it thing) or if it would disable whenever wg-manager reloaded.
But I had put mine on Enable.

Anyways as best as I can tell Flow Cache seems to disable itself shortly after the needles move on the QoS - WAN/LAN Bandwidth Monitor
but it stays disabled, afterwards.
So for the time being... I just stay the hell outta there LOL

 

[AntonK]

For the first time that I can remember, something on my network has changed with the installation of a new firmware. Since Since installing 386.5, my Echo Studio drops Spotify, for a second, every so often.

I did a full reset of both my router and my Echo device, but no change.

Weird
Anton