Release Asuswrt-Merlin 388.2 is now available for select models

[Makaveli]

Both, At internal speedtest have less loss of speed.
Reverting to 388.2 solved in both cases.

I don't see any reduction in speed when I test on my end.

 

[Bogey]

We're both running same dnsmsq, GPL versions but different models. When you said you factory resetted the router, did you install diversion from scratch too? Can you try installing diversion lite instead of using pixelserve?

Ok, the results of testing Diversion lite with logging on: disconnected devices. With logging off everything works. So it’s not pixelsserve. Once again confirmed that it’s the logging of Diversion that triggers that devices can’t connect and a restart of dnsmasq fixes it.

 

[bluepoint]

Ok, the results of testing Diversion lite with logging on: disconnected devices. With logging off everything works. So it’s not pixelsserve. Once again confirmed that it’s the logging of Diversion that triggers that devices can’t connect and a restart of dnsmasq fixes it.

If diversion logging works for me why does it not for you? Your USB stick used to have UIdivstat installed I'm thinking there might be a corrupted data, is it easy for you to start from scratch with your addons? Uninstall all addons so it removes remnant files from jffs then format the usb. Install diversion again with logging then observe if your devices disconnects. Do you have static IP's? If yes, how many?

 

[saison2023]

If diversion logging works for me why does it not for you? Your USB stick used to have UIdivstat installed I'm thinking there might be a corrupted data, is it easy for you to start from scratch with your addons? Uninstall all addons so it removes remnant files from jffs then format the usb. Install diversion again with logging then observe if your devices disconnects. Do you have static IP's? If yes, how many?

I have to admit that I have a long list of assigned IP to my devices, doing MAC/IP/hostname mapping in "LAN - DHCP Server" (around 30 mapping)
I may try to "reset" (or remove) the reservation for almost all the devices and give a try to confirm if your theory about having a long static IP list is true or not.

 

[Bogey]

If diversion logging works for me why does it not for you? Your USB stick used to have UIdivstat installed I'm thinking there might be a corrupted data, is it easy for you to start from scratch with your addons? Uninstall all addons so it removes remnant files from jffs then format the usb. Install diversion again with logging then observe if your devices disconnects. Do you have static IP's? If yes, how many?

Not only me, there are more.
USB has been renewed too. Jffs been formatted too. Thing is when we go flash dirty(!) to 388.1 it all works again.

I appreciate your input, I really do, but we going full circle. We (not just me) really covered most if not all probable causes. (You have to go back through this thread to find them as they’re scattered throughout).

But, we’re in touch with the developer dnsmasq. Though it requires more than just our findings to get to the bottom of the issue.

 

[Bogey]

I have to admit that I have a long list of assigned IP to my devices, doing MAC/IP/hostname mapping in "LAN - DHCP Server" (around 30 mapping)
I may try to "reset" (or remove) the reservation for almost all the devices and give a try to confirm if your theory about having a long static IP list is true or not.

I tried it to with a basic config after a factory reset and only Diversion. IP assignment was not part of that and still I had disconnects.

 

[saison2023]

I tried it to with a basic config after a factory reset and only Diversion. IP assignment was not part of that and still I had disconnects.

Thanks, you saved me a lot of time, especially if the result is a not positive one.

 

[bluepoint]

But, we’re in touch with the developer dnsmasq. Though it requires more than just our findings to get to the bottom of the issue.

That's good, taught you might want to explore to pinpoint to the develeloper the direction to squash the bug. It helps them to track the problem. Anyway, good luck.

 

[Bogey]

That's good, taught you might want to explore to pinpoint to the develeloper the direction to squash the bug. It helps them to track the problem. Anyway, good luck.

Yes I do and I appreciate the input, though I think we’re in need for some debug versions that log extra stuff. We’ve narrowed it down to these last two “variables”.

 

[plexion]

I am seeing some weird CPU usage issues after upgrading to 388 from 386.7_2 on my Asus RT-AX56U.
I have tried 388.1, 388.2 Beta and now 388.2, all resulting in the same outcome.

Issues:
1. Only 3 out of 4 CPU cores of the router is utilized. 4th core is always idle.
2. Higher than usual CPU usage on 388 firmwares compared to 386.7_2 constantly, even on idle. The "sysstate" process using up quite a lot of CPU time.


CPU Usage Graph on 388.2

View attachment 49449


CPU Usage Graph on 386.7_2

View attachment 49450


Top Command Output on 388.2

View attachment 49453


Top Command Output on 386.7_2

View attachment 49454

I have tried:
1. Dirty flashing from 386.7_2 to the 388 series firmwares.
2. Factory reset on 386.7_2, then flashing to 388 series firmwares.
3. Factory reset on 388 series firmwares after upgrading from 386.7_2.

All outcome are the same as above.

Going back to 386.7_2 also always resolves the problem.

Appreciate if anyone can help to shed some light on the behaviours of the router above.
Thanks!

I have tried upgrading to 388.2_2 today, both dirty and clean flashes.
Unfortunately the issue still remains.

Downgraded to 386.7_2 and everything is good again.

@RMerlin will you be able to advise please? Thanks.

 

[BloodFX2021]

Hi is anyone getting start stubby spam in system log with 388.2_2 on ax88u ?

May 17 09:09:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:09:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:10:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:10:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:10:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:10:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:11:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:11:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:11:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:11:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:12:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:12:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:12:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:12:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:13:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:13:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:13:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:13:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:14:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:14:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:14:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:14:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:15:19 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:15:19 custom_script: Running /jffs/scripts/service-event (args: start stubby)
May 17 09:15:49 rc_service: watchdog 2134:notify_rc start_stubby
May 17 09:15:49 custom_script: Running /jffs/scripts/service-event (args: start stubby)

 

[prosperot]

Both, At internal speedtest have less loss of speed.
Reverting to 388.2 solved in both cases.

I've had the same problems after a dirty flash from 386.2 to 386.2_2. A clean install solved my speed problems.

 

[slackjaw99]

Hello, good afternoon.

I have an RT-AXE7800 and I read here that the Asuswrt-Merlin version is compatible with all routers with Wi-Fi 6.
So, could I install this version?
Or when you mean that it is compatible with all routers with Wi-Fi 6, are you referring only to those that currently appear on your website?

Thank you so much

That model is a dud IMO. I doubt it will ever get Merlin support. Bought 2 of em, and started the return process less than 24 hours after delivery cuz you can't turn wifi 6 mode off for any of the lower bands like you can with every other ASUS 6E. 600$ and all we could connect were two brand new Pixel phones out of 50+ computers and IoT devices. The ET8 has the same specs plus the ability to toggle wifi 6 off for 5g/2.4g if desired or needed for 1/2 the price.

 

[RMerlin]

@RMerlin will you be able to advise please? Thanks.

All cores do get used, you can see them with top. Core usage will vary by affinity, some processes like OpenVPN are locked to specific cores.

If something has high CPU usage then you need to track it down using top.

 

[GHammer]

May 16 17:07:56 AsusRouter dnsmasq[3393603]: started, version 2.89 cachesize 1500

Shows you are using the buggy version that was replaced (previous version) in the newest load.
From the readme in 388.2.2
FIXED: dnsmasq may crash if no DNS server is configured
(fix backported from dnsmasq upstream)

 

[RMerlin]

Shows you are using the buggy version that was replaced (previous version) in the newest load.
From the readme in 388.2.2
FIXED: dnsmasq may crash if no DNS server is configured
(fix backported from dnsmasq upstream)

388.2_2 is still using dnsmasq 2.89, I only backported a fix for a crash that can occur if there are no servers configured. The dnsmasq version is unchanged.

 

[saison2023]

This morning, I did a Factory reset (FW 388.2_2) using WPS button method provided by Asus

I only made these changes:

• Wifi (during first setup after factory reset)
• Guest Network 2,4GHz (for IoT and appliances)
• LAN (change router IP)
• Used a totally different USB stick
• Activate JFFS
• Change DHCP range
• WAN - Virtual Server / Port Forwarding
• WAN - DDNS
• In AMTM:
  • Format USB, Create swap and configure disk checker
  • Install Diversion Standard (pixelserv-tls), medium blocking list, activate YouTube blocking
  • Install scMerlin

And after around 6 hours, I got an error:

May 17 15:14:27 roamast: sta[6E:BF:55:39:7B:CF] on ap[7C:10:C9:B4:CB:C4], rcpi is 84 and rssi is -68
May 17 15:23:04 rc_service: httpds 1717:notify_rc start_diversion
May 17 15:23:04 custom_script: Running /jffs/scripts/service-event (args: start diversion)
May 17 15:26:06 kernel: potentially unexpected fatal signal 11.
May 17 15:26:06 kernel: CPU: 3 PID: 1717 Comm: httpds Tainted: P O 4.1.52 #2
May 17 15:26:06 kernel: Hardware name: Broadcom-v8A (DT)
May 17 15:26:06 kernel: task: ffffffc03dc04b00 ti: ffffffc02cf8c000 task.ti: ffffffc02cf8c000
May 17 15:26:06 kernel: PC is at 0x4d4c0
May 17 15:26:06 kernel: LR is at 0x4d484
May 17 15:26:06 kernel: pc : [<000000000004d4c0>] lr : [<000000000004d484>] pstate: 80070010
May 17 15:26:06 kernel: sp : 00000000ff827f10
May 17 15:26:06 kernel: x12: 00000000000c42c0
May 17 15:26:06 kernel: x11: 0000000000000001 x10: 00000000ff8281bc
May 17 15:26:06 kernel: x9 : 0000000000803950 x8 : 0000000000000000
May 17 15:26:06 kernel: x7 : 00000000ff828034 x6 : 00000000ff8281bf
May 17 15:26:06 kernel: x5 : 00000000f73d3ab8 x4 : 0000000000000000
May 17 15:26:06 kernel: x3 : 00000000000a100a x2 : 00000000ffa8a600
May 17 15:26:06 kernel: x1 : 0000000000000211 x0 : 0000000000000000

This test was to validate @Bluepoint theory about Static assignment in DHCP and/or related (like MAC reservation,..)
I did nothing like that (see my list above), and there was no trouble doing reservation and static assignment with previous dnsmasq version with firmware 388.1 and previous.
This seems to confirm the trouble with dnsmasq but may remove one possible clue.
Hope this will help to narrow down the trouble.

 

[Stephen Harrington]

Shows you are using the buggy version that was replaced (previous version) in the newest load.
From the readme in 388.2.2
FIXED: dnsmasq may crash if no DNS server is configured

See @RMerlin's reply above re: dnsmasq 2.89, but in any case I do definitely have DNS servers configured in "WAN | WAN DNS Setting | DNS Server" section if that is what this "FIXED" note is referring to? Probably slightly confused (in my mind) by the fact I run @SomeWhereOverTheRainBow's AMTM AdGuardHome add-on which I believe takes over all of that? Definitely getting into stuff that I don't pretend to understand

 

[bluepoint]

This morning, I did a Factory reset (FW 388.2_2) using WPS button method provided by Asus

I only made these changes:

• Wifi (during first setup after factory reset)
• Guest Network 2,4GHz (for IoT and appliances)
• LAN (change router IP)
• Used a totally different USB stick
• Activate JFFS
• Change DHCP range
• WAN - Virtual Server / Port Forwarding
• WAN - DDNS
• In AMTM:
  • Format USB, Create swap and configure disk checker
  • Install Diversion Standard (pixelserv-tls), medium blocking list, activate YouTube blocking
  • Install scMerlin

And after around 6 hours, I got an error:


This test was to validate @Bluepoint theory about Static assignment in DHCP and/or related (like MAC reservation,..)
I did nothing like that (see my list above), and there was no trouble doing reservation and static assignment with previous dnsmasq version with firmware 388.1 and previous.
This seems to confirm the trouble with dnsmasq but may remove one possible clue.
Hope this will help to narrow down the trouble.

This isn't dnsmasq crash. It's related to httpds probably caused by pixelserve. Try diversion lite and see what happens.

 

[Bogey]

This isn't dnsmasq crash. It's related to httpds probably caused by pixelserve. Try diversion lite and see what happens.

@saison2023 had the same results I did. Regarding Diversion lite/pixelserve… Didn’t I already reported that back to you? Because that’s what I have already confirmed: it’s NOT pixelserve. It’s the logging of Diversion (as that’s switched on by default) in combination with dnsmasq 2.89