Release Asuswrt-Merlin 3004.388.10 is now available

[dbell]

A new stock firmware was posted yesterday:

 

[ColinTaylor]

A new stock firmware was posted yesterday:
View attachment 68490

And how is that relevant to this thread?

 

[dbell]

And how is that relevant to this thread?

Well we're talking about instability in Merlin FW 10_0, and he just determined it is in closed source, but Asus clearly recognized an issue given the first two comments in the release notes and it's only 6 wks after their last release which have been coming out in at least several months increments recently. That is all evidence that what we are experiencing is not limited to mobile app, or HA integrations, but rather is a FW problem itself.

So folks have a choice to stick with 9_2, or try the new stock firmware instead and see if it resolves the issues, or limp along with 10_0 until another Merlin update can roll in a suitable GPL.

Is that relevant enough for you ?

 

[ColinTaylor]

Is that relevant enough for you ?

Then you could have just said "try the stock firmware" rather than posting the changelog for one specific model without explanation.

 

[dbell]

Then you could have just said "try the stock firmware" rather than posting the changelog for one specific model without explanation.

Ah ok got it - you just didn't like the way I communicated that Asus published a very recent firmware update that specifically targets stability problems.

"Try the stock firmware" has even less context and relevance, but I'll bear your advice in mind for the future.

 

[Clark Griswald]

Could you please clarify your concern here, as to divergence from the topic?

With all due respect, if this particular FW causes an issue with any system that integrates, communicates or interconnects with it, whether that be internal or external to the Routers “OS” for want of better terminology, then surely by definition it is on topic? That it happens to be HA in this instance is irrelevant.

Granted, commenting on the pros and cons of HA or its integrations is a divergence; but in my case above I was suggesting an interim solution for those who wanted to use the new FW and avoid the issues associated with HA that came with it, not to extol the virtues of HA itself. I do not think this unreasonable, in this particular instance, but I am open to hearing your concern.

Please state issues that directly effect RMerlin's FW and HA with any additional help (logs, screen shots, etc) in addressing the issue. Since HA is not part of Asus nor RMerlin FW I suggested you possibly open your issue in asuswrt-merlin addons Many of these technical threads get lengthy and finding solutions is difficult. Thus your questions and solutions could be easily found and addressed, instead of currently buried on page #10.
Or just run HA on a RPi and let the router do what it was designed.

 

[jksmurf]

Please state issues that directly effect RMerlin's FW and HA with any additional help (logs, screen shots, etc) in addressing the issue

I do not personally have these issues but feel free to check this thread for folks describing their issues with HA as a direct result of this FW, along with Merlin’s conclusion that the problem rests with Asus closed source part of the FW and is not something he can address. If you refer HA posters to that conclusion rather than requesting they totally desist from posting about HA issues related to this firmware, that would be a positive addition, in my view.

Clearly there are e.g. here folks that have issues with the WebGUI (e.g.) on this firmware when used with a HA integration; issues that they did not experience with the previous FW revision.

I do not think it outwith the bounds of reason for them to raise that here.

Or just run HA on a RPi and let the router do what it was designed

Your recommendation that they simply run HA on a RPi suggests somewhat of a misunderstanding of the issue, as regardless of the device HA is run on, if the HA Integration refers to the router in the network, the one with this FW on it, the issue will still arise.

Ultimately I appreciate your concerns regarding posts diverging from the FW posted and I also find this detracts from nailing down collective, repeatable issues with the FW, but in this instance, sir, I think the HA issues are and remain, relevant.

 

[aex.perez]

Confused

Had lots of headaches with 3004.388.10, reverting back to beta1 to stabilize with just about every service failing and restarting. Then having to get on the develop track of the scribe add it to get logging and have my custom filters working again. Even had the Node not passing traffic with BACKUPMON failing to find the NAS, couldn't even ping the Router via SSH on the Nodes, though they could still be managed via the Router through the GUI, a reboot of each fixed that. Still with all that I was was not able to use the App, let alone the App and the GUI simultaneously and didn't risk even trying.

Then last night, after a few days of stability, I took a the risk. Downloaded the App on the iPhone and set it up and it was able to connect, login and actually report on active connections on the Router and on each Node

So, file under no risk no reward, took the next step. Tried this morning, being in the GUI and using the App at the same time. Before it wouldn't work and it would cause a cascade of failures with Router processes, firewall, logging, ACS, channel changes amongst others.

But this morning, both were working like they did under 3004.388.9.2 with the Router and Nodes remaining stable, though the log entries are different now.

Oct 21 08:43:43 Router HTTPD: [LOGIN][http][Web] success (192.168.1.99)
Oct 21 08:45:45 Router HTTPD: [LOGIN][https][APP] success (192.168.1.172)

Continues to work Not sure how or why but I'll take the WIN!

Well it mostly all works now.

Have a scheduled job that does not appear to run, or more acurately actually does not do what its supposed to do when it's scheduled to do it. But works fine from the command line (and did under 9.2), and does run at it's scheduled time
The schedule:
0 4 * * * /bin/sh /jffs/scripts/clearcache.sh #ClearCache#

The script:
#!/bin/bash
echo 3 > /proc/sys/vm/drop_caches && logger 'Ram-cache and Swap Cleared'

As far as the other random issues that that come about from using the APP or with HTTP/HTTPS or start after a few days of running fine. The AX88 Router has been working fine, no errors in the log. Thought since I also run Merlin on the nodes (quite successfully for quite some time) that issue has moved from the Router and onto both AX86 nodes, causing them to crash, though they do restart. As they complan "kernel: CPU: 0 PID: 13273 Comm: httpds Tainted: P O 4.1.52 #2".

At the height of issues with my AX88 with the release of 3004.388.10, with all sorts of scripts and services failing, causing me to roll back to Alpha1 for stability and abandon the APP. Well, just sharing mostly as "the APP so not supported" with Merlin FW, and though I'm only using it for visibility and maybe for a clue into what was/is going on. Though @RMerlin did post on the HTTP service and logging being closed source to clear it up, just adding this for any additional clarity this may provide, if any.

Logs from the nodes and the crash attached.

 

[LaMotta]

Let me share a not expert kind but a regular user comment.
After some issues already reported here, additionally I have one as well.
I did a clean setup at 3004.388.10 after 3004.388.9_2 on an AX86S as a main router.
There are two additional AiMesh nodes, an AX55 and an AX57 Go.
AX86S reboots "periodically", more precisely randomly, since 3004.388.10.
I did a dirty revert to 3004.388.9_2, works fine again as before.

(Praying for 3004.388.11 in the next some month. )

 

[Wishmaster1965]

Did a dirty Upgrade on router and mesh node.
Seeing an issue with 0 clients showing on aimesh screen on either node and I have lots of clients connected to lan and wifi.

 

[zemerdon]

Upgraded like normal through the firmwares and am now receiving the following error from dmesg...
(web access is being interrupted and it cannot complete tasks, refuses connection etc, i might get a few clicks in before it crashes)

potentially unexpected fatal signal 11.

CPU: 0 PID: 3807 Comm: httpd Tainted: P           O    4.1.51 #2
Hardware name: Broadcom-v8A (DT)
task: ffffffc02bcd1540 ti: ffffffc02f0e4000 task.ti: ffffffc02f0e4000
PC is at 0xf6b9c72e
LR is at 0xf6b68858
pc : [<00000000f6b9c72e>] lr : [<00000000f6b68858>] pstate: 200d0030
sp : 00000000ffd06a28
x12: 00000000ffffffff
x11: 00000000ffd06f24 x10: 0000000000000000
x9 : 00000000f6c627a8 x8 : 00000000f6c64000
x7 : 00000000f74061b0 x6 : 000000000000002d
x5 : 00000000000b651c x4 : 0000000000000004
x3 : 0000000000000000 x2 : 0000000000000001
x1 : 00000000000001c0 x0 : 00000000000001c4

zemerdon@RT-AX88U-CA30:/tmp/home/root#

I can post the full dmesg if required. Anyways, love the firmware, going to downgrade a version and see if it fixes it...

UPDATE: rolled back to 'RT-AX88U_3004_388.9_2_ubi.w', no issues.

 

[The PyroPath]

Please state issues that directly effect RMerlin's FW and HA with any additional help (logs, screen shots, etc) in addressing the issue. Since HA is not part of Asus nor RMerlin FW I suggested you possibly open your issue in asuswrt-merlin addons Many of these technical threads get lengthy and finding solutions is difficult. Thus your questions and solutions could be easily found and addressed, instead of currently buried on page #10.
Or just run HA on a RPi and let the router do what it was designed.

So the thing for me is that the App and HA issues are "among others". As I stated I was unable to log in, was treated on a https screen I've never seen before, while I was connecting through 8443 and have been doing for years now. When I connected the dots that the App and HA integration were causing issues, I disabled them and have also stated that, multiple times now. But still I'm having issues with the firmware with the GUI not working, getting stuck in a logging in loop, when I manage to log in, the GUI loads very slow, pages are corrupted, showing features my router/Aimesh node does not possess. My log is riddled with CFG-errors, channel changes, connecting and disconnecting devices. Aimesh has turned into a aimless with nodes disconnecting, not using the correct backhaul, (2.4 vs 5 GHz) or not connecting in general. While happily browsing one moment, WAN is disconnected the next, resulting in devices complaining that the network has no internet connection, which isn't showing up in the log and the network page still proudly shows that it's connected to the internet just fine. Instant Guard has stopped working properly. I can still connect just fine, but once connected, I can't browse pages anymore unless I use the direct IP address. And I could go on. I've tried hard resets, reflashes, stock settings, a roll back, but the issues stay. I would be lying if I said I'm not getting frustrated with people just rehashing comments that the App and HA are not supported, because the issues stay, even when I disabled the app and HA integration.

It seems rather random. Some people have no problems, some have more problems and the last (like me) have a crippled network. It's so bad I shelled out the cash to get a new router and AP's. Unfortunately that is not all sunshine and roses either, but that is something else.

6 weeks ago we had twins, so I'm quite preoccupied with caring for them. I couldn't continue spending every spare time with fixing the Wi-Fi. But that there is more going on than just the App connection is clear to me.

 

[zemerdon]

So the thing for me is that the App and HA issues are "among others". As I stated I was unable to log in, was treated on a https screen I've never seen before, while I was connecting through 8443 and have been doing for years now. When I connected the dots that the App and HA integration were causing issues, I disabled them and have also stated that, multiple times now. But still I'm having issues with the firmware with the GUI not working, getting stuck in a logging in loop, when I manage to log in, the GUI loads very slow, pages are corrupted, showing features my router/Aimesh node does not possess. My log is riddled with CFG-errors, channel changes, connecting and disconnecting devices. Aimesh has turned into a aimless with nodes disconnecting, not using the correct backhaul, (2.4 vs 5 GHz) or not connecting in general. While happily browsing one moment, WAN is disconnected the next, resulting in devices complaining that the network has no internet connection, which isn't showing up in the log and the network page still proudly shows that it's connected to the internet just fine. Instant Guard has stopped working properly. I can still connect just fine, but once connected, I can't browse pages anymore unless I use the direct IP address. And I could go on. I've tried hard resets, reflashes, stock settings, a roll back, but the issues stay. I would be lying if I said I'm not getting frustrated with people just rehashing comments that the App and HA are not supported, because the issues stay, even when I disabled the app and HA integration.

It seems rather random. Some people have no problems, some have more problems and the last (like me) have a crippled network. It's so bad I shelled out the cash to get a new router and AP's. Unfortunately that is not all sunshine and roses either, but that is something else.

6 weeks ago we had twins, so I'm quite preoccupied with caring for them. I couldn't continue spending every spare time with fixing the Wi-Fi. But that there is more going on than just the App connection is clear to me.

AX88U + XD6S nodes, flawless for me. 410m square house.

 

[Pergola Fabio]

i have a brand new rt-ax88u, upgraded instantly to this firmware, did a factory reset
running a wheek stable, they suddenly no wifi anymore, i connected to lan port of the router, no dhcp ....
only thing left todo was , restart my router, i see messages like those in my logs

Oct 25 14:22:18 kernel: FPM Pool 1: invalid token 0x22017000 freed
Oct 25 14:22:18 kernel: FPM Pool 1: ISR timer is enabled. There could be multiple occurrences of the reported issue
Oct 25 14:22:22 kernel: FPM Pool 1: invalid token 0x245b0000 freed
Oct 25 14:22:22 kernel: FPM Pool 1: ISR timer is enabled. There could be multiple occurrences of the reported issue


Oct 25 14:23:25 kernel: CFG80211-ERROR) wl_cfg80211_change_station : WLC_SCB_AUTHORIZE sta_flags_mask not set


Oct 25 14:22:58 kernel: ERR: rdpa_cpu_tx_port_enet_lan#213: rdpa_cpu_tx_port_enet_lan failed. rdd_rc=120 tx_rdd_error_count=1

anyone else seeing this too ?

 

[Pergola Fabio]

@Khadanja @joegreat

I see you have those rdpa errors too?
What happens in your case? Also on ax88u ?
Mine completely freezed... I noticed the wifi was gone, also connection with lan didn't work, I had to reboot my router...
Did yours router restore automatically after a while? Because if I'm not home, I can't reboot offcourse... Did you resolve the issue? Maybe downgraded ?

Thnx

 

[bennor]

anyone else seeing this too ?

There is a discussion started March of last year where a user posted similar system log messages. May want to read through it, if you haven't already, to see if there are any suggestions that may apply.

Router trouble RT-AX88U

My router is Asus RT-AX88U, firmware version 3004.388.6. The 8 port switch is used as follows: - port 1: desktop - port 2: printer - port 3: NAS - port 4: powerline adapter - port 5: Raspberry Pi running PiHole DHCP is enabled on router. PiHole works only as DNS (DHCP disabled). The router...

www.snbforums.com

 

[dbell]

So the thing for me is that the App and HA issues are "among others". As I stated I was unable to log in, was treated on a https screen I've never seen before, while I was connecting through 8443 and have been doing for years now. When I connected the dots that the App and HA integration were causing issues, I disabled them and have also stated that, multiple times now. But still I'm having issues with the firmware with the GUI not working, getting stuck in a logging in loop, when I manage to log in, the GUI loads very slow, pages are corrupted, showing features my router/Aimesh node does not possess. My log is riddled with CFG-errors, channel changes, connecting and disconnecting devices. Aimesh has turned into a aimless with nodes disconnecting, not using the correct backhaul, (2.4 vs 5 GHz) or not connecting in general. While happily browsing one moment, WAN is disconnected the next, resulting in devices complaining that the network has no internet connection, which isn't showing up in the log and the network page still proudly shows that it's connected to the internet just fine. Instant Guard has stopped working properly. I can still connect just fine, but once connected, I can't browse pages anymore unless I use the direct IP address. And I could go on. I've tried hard resets, reflashes, stock settings, a roll back, but the issues stay. I would be lying if I said I'm not getting frustrated with people just rehashing comments that the App and HA are not supported, because the issues stay, even when I disabled the app and HA integration.

It seems rather random. Some people have no problems, some have more problems and the last (like me) have a crippled network. It's so bad I shelled out the cash to get a new router and AP's. Unfortunately that is not all sunshine and roses either, but that is something else.

6 weeks ago we had twins, so I'm quite preoccupied with caring for them. I couldn't continue spending every spare time with fixing the Wi-Fi. But that there is more going on than just the App connection is clear to me.

QED - just one example symptom.

The only log entries are these after refreshing the page:
Oct 27 12:05:18 HTTPD: [LOGIN][https][Web] success (192.168.244.1)
Oct 27 12:05:26 HTTPD: [LOGIN][https][APP] success (192.168.241.1)

244.1 is the machine from which I took this screenshot, 241.1 is a QNAP NAS running home assistant container.

I am coming more and more to believe this is related to invalidation of authentication state / tokens by having more than one entity connecting, authenticating and using the API of the router and is likely a side-effect of a change in authentication flow mentioned in ASUS release notes (but there is no detail about how it changed):

 

[bg604]

I have the BE98 Pro, when I go to the WAN DNS entry. Under Fast DNS, Google is only listed. When I manually assign CloudFlare 1111 and 1001, it picks it up and lists it as:

Filter Mode: Fast DNS
Service Name: Cloudflare
DNS Server: 1.1.1.1, 1.0.0.1

In other words, CloudFlare is missing from the Fast DNS option.

 

[visortgw]

I have the BE98 Pro, when I go to the WAN DNS entry. Under Fast DNS, Google is only listed. When I manually assign CloudFlare 1111 and 1001, it picks it up and lists it as:

Filter Mode: Fast DNS
Service Name: Cloudflare
DNS Server: 1.1.1.1, 1.0.0.1

In other words, CloudFlare is missing from the Fast DNS option.

This appears to be consistent across the Merlin builds — only a handful of DNS providers in the configuration file are enabled in the GUI.

That being said, this is not the proper thread to bring up issues with GT-BE98 Pro firmware as you clearly are not running Asuswrt-Merlin 3004.388.10 firmware on your router.

 

[RMerlin]

Asus fixed the httpd crashing issue. I haven't had the chance to test it myself yet, we are working out a way for me to get the fix since it resides inside a closed source component.