An Observation- 386 Code, AX86U, and WiFi Failure

[GHammer]

I have observed that my AX86U loses the ability to pass traffic via WiFi once per three day period. I say once in a three day period because that is as long as I've seen things stay up while one day is the shortest period.

Doesn't matter if I run official or Merlin, once in a three day period, phones and other WiFi devices will all lose the ability to reach anywhere on the Internet. This affects both 2.4 and 5. Radios are up, I can see connections in the router, devices show good signal, etc. Just no Internet is to be reached.

No worthwhile logs are present. Wired devices have the same Internet access as always. A quick reboot of the router restores the Internet for WiFi based devices.

With no definite cause, no certain method to reproduce, and no logs, I don't really expect to solve this. Just info for others who may suffer from the same symptoms.

 

[Rassal]

Haven't seen this on my RT-AX86U Yet, but i never ran stock firmware, been using it with Merlin in beta since day 1 i bought it.

Longest i ran the router without rebooting is around 10 days (mostly because of firmware updates).

Never experienced any problem with 2.4 or 5 radio, and i have to say i am free of Weather radar in my Area, so i am lucky to run 5 band in 160mhz without problem.

I will keep an eye open, but as i said, 10 days without rebooting is my longest... it all depends on new firmware releases...

 

[L&LD]

The symptoms you describe are always 'cured' for my customers when I follow the suggestions below as needed.

Best Practice Update/Setup Router/AiMesh Node(s) 2021

Flash the firmware you want to use (RMerlin 386.1_2 release final, today).

Reset the router to factory defaults afterward and then minimally and manually configuring the router to secure it and connect to your ISP.

Do not import a saved backup config file or 'blindly' use old settings/options or scripts that worked on previous firmware or router.

 

[thiggins]

@L&LD Can you describe what is causing the problem the OP is reporting and why your suggestions fixes it?

 

[L&LD]

I don't, 'know' what is causing the problem, specifically. But misconfiguration somewhere along the line for the wireless environment is inferred. Particularly when following the 'best practice' suggestions fixes things (for my customers and others here on this forum).

Not only for this specific model, but for others, too.

As I have repeated many times previously if the steps suggested are followed (precisely, in order, and fully) and the symptoms persist, that would indicate a hardware issue to me.

 

[thiggins]

No offense intended. But when you can't identify the cause of a problem, you can see why some question the value of going through all the steps you describe as a fix.

 

[L&LD]

I didn't identify it specifically, but it is inferred.

No offense is taken.

The value is getting a stable and reliable network once again.

 

[GHammer]

The symptoms you describe are always 'cured' for my customers when I follow the suggestions below as needed.

Best Practice Update/Setup Router/AiMesh Node(s) 2021

Flash the firmware you want to use (RMerlin 386.1_2 release final, today).

Reset the router to factory defaults afterward and then minimally and manually configuring the router to secure it and connect to your ISP.

Do not import a saved backup config file or 'blindly' use old settings/options or scripts that worked on previous firmware or router.

Yeah, well, the router had been wiped when I installed 386.1, same issue.
I understand that you are a big proponent of wipe it down and start again, but it doesn't apply to each and every issue and is tiring to read. again, and again, and again.

 

[GHammer]

The value is getting a stable and reliable network once again.

Again, if there was a problem that indicated a clean start is in order.
In this case, the suggestion is the same as 'Have you turned your computer off and on"

 

[L&LD]

The router needs to be reset after flashing the firmware you want to try/test.

I've never stated it should be done for every issue either.

 

[GHammer]

The router needs to be reset after flashing the firmware you want to try/test.

I've never stated it should be done for every issue either.

Again, I know how to do it and have.
The why is the thing.

 

[L&LD]

These days, even turning devices on/off may not be enough. For Windows, holding down the Shift Key and then clicking 'Shut down' (and then leaving the device off for a minute or two), is different than just clicking Shut down or hitting the power button. A reboot may be as effective as the Shift Shut down method too, but not in all cases either, depending on what the problem is.

I don't claim to know a specific reason for the 'why'. I just know it works. Not from my head. But from the experience of the past members here, who helped me get to where I am today with RMerlin powered Asus routers. Their suggestions starting from 8 years ago and earlier are what I base my 'best practice' steps on.

If a quick search doesn't find a resolution to any issues someone is facing, continuing that search and testing random possible fixes more than likely means more time and energy wasted while suffering from the issues, still.

A full/proper reset to get a network up and running is quicker and will either prove to be the correct answer, point to a hardware issue, or suggest a possible bug in the firmware. All those situations are a good enough 'why' for me and my customers.

To get a network 'up' is less than 15 minutes, max, and as little as 5 minutes, for more straightforward networks. Putting off a reset because of time is not an excuse. Asking for 'why', is.

The routers today are doing so much more than routers of even just a few years ago. Users are expecting more too, from their routers/networks. With today's routers needing to bridge up to 6+ generations of Wi-Fi in a single network environment, along with firmware that is more complex, more feature-rich, and with more interdependencies, internally and externally between the open-source sections and the closed source blobs and the features/options enabled (or not). And all this in just the last few years I've been actively following them.

The SoC used in the hardware isn't really a full-on 'computer' or straight 'Linux'. Rather, something in-between.

All of the above and more are 'why' a reset is usually the fastest way back to a stable router and network.

 

[dave14305]

As an amateur developer, I suggest that rushing to a factory reset may help that one user, but destroys "evidence" that might help the developer get to the root cause to be able to fix it permanently for everyone else. No one cares about "everyone else" when their router isn't working at all, so it's more an ideal than a likely outcome.

It's more expedient to start fresh sometimes, but the underlying issues still linger in the wild, unresolved. Our corporate help desk will spend a few minutes troubleshooting someone's PC issue, then abruptly suggest the end user "reimage their laptop" as a "solution." It probably works, but it's often overkill for the magnitude of the original problem. But no one wants to invest the time to get to the bottom of the real issue. Time is money.

 

[L&LD]

Depending on the specific 'issues' being faced by a particular user, and that user's own threshold criteria, 'rushing' to factory reset is a very fluid timeframe.

Customers call for 'emergencies' and are okay if I show up in a week. Others know they have unimportant issues (to the daily network use) but will suggest/demand 'today?'.

This alludes to that 'quick search' for a resolution (above). One person may want to try to know the 'why' and spend what someone else may deem to be an inordinate amount of time to implement a 'fix'.

When customers ask me for the 'why', I tell them it will cost more than the fix.

 

[J.I]

I have observed that my AX86U loses the ability to pass traffic via WiFi once per three day period. I say once in a three day period because that is as long as I've seen things stay up while one day is the shortest period.

Doesn't matter if I run official or Merlin, once in a three day period, phones and other WiFi devices will all lose the ability to reach anywhere on the Internet. This affects both 2.4 and 5. Radios are up, I can see connections in the router, devices show good signal, etc. Just no Internet is to be reached.

No worthwhile logs are present. Wired devices have the same Internet access as always. A quick reboot of the router restores the Internet for WiFi based devices.

With no definite cause, no certain method to reproduce, and no logs, I don't really expect to solve this. Just info for others who may suffer from the same symptoms.

I experienced that once with my AX86U which I bought two weeks ago and right away loaded Merlin. Not sure if wired devices still had access to the network as when I connected my laptop thru ethernet it could not obtain IP from DHCP of that router. By any chance do you run Guess network 1 or had run it in the past on that router?

 

[GHammer]

Depending on the specific 'issues' being faced by a particular user, and that user's own threshold criteria, 'rushing' to factory reset is a very fluid timeframe.

Customers call for 'emergencies' and are okay if I show up in a week. Others know they have unimportant issues (to the daily network use) but will suggest/demand 'today?'.

This alludes to that 'quick search' for a resolution (above). One person may want to try to know the 'why' and spend what someone else may deem to be an inordinate amount of time to implement a 'fix'.

When customers ask me for the 'why', I tell them it will cost more than the fix.

Only if your actions actually result in a fix. Which, in this case, it would not.
I don't think I'd hire someone who doesn't really correct my issues.

 

[GHammer]

I experienced that once with my AX86U which I bought two weeks ago and right away loaded Merlin. Not sure if wired devices still had access to the network as when I connected my laptop thru ethernet it could not obtain IP from DHCP of that router. By any chance do you run Guess network 1 or had run it in the past on that router?

No, no Guest network on the router. Except for YazDHCP, a plain vanilla install.

 

[L&LD]

@GHammer, of course, my actions result in a fix. What case are you talking about?

I don't charge by the hour or by the mile, I charge by the services I provide and the resolution of a problem for my customers. Customers who refer me to the next paying job, BTW.

 

[Sky1111]

Folks, I have been observing this discussion and IMHO the current argument is not very productive. L&LD is sharing what works for him, and of course there is a difference between workaround (such as full reset/erase/etc.) and rootcausing/debugging and properly fixing the issue.
Sometimes the workaround becomes the solution - anyone who works with software would understand - as a proper solution may simply not practical (for example, some issues require silicon or platform change, so that can only help the next gen, but not the current user).

The real question here: how do we narrow down and document the issues so they become actionable for ASUS FW team or Merlin?
From work experience, what helps is a creation of a standard bug reporting template that can be filled and additional history addeded (for example, the same issue is seen on AX86, AC86, etc., not reported on AX88 - just for example)

 

[Timo]

I had more or less the same problem.

After a day+ the dns resolving would crawl (10 seconds or more) and stop altogether after 1-3 days.
But established connections would work at normal speed. Clearly a dns problem.
As expected configuring clients to use an alternative DNS server solved that.

After a lot of troubleshooting (resetting and reconfiguring each time) I found one of the add-ons was causing the behavior. Removing the webgui stats of diversion and skynet was the fix for me.
I didn't bother to find out which one was the culprit.