My experience with the RT-AC86U

[Viktor Jaep]

This is strange. The router is quite old. No one noticed something is off? It was a popular router few years back.

I remember running one of Martineau's failover scripts back then, and it would hang on occasion... It happens so far and few between... perhaps people just wrote it off as normal?

 

[SomeWhereOverTheRainBow]

I remember running one of Martineau's failover scripts back then, and it would hang on occasion... It happens so far and few between... perhaps people just wrote it off as normal?

cycling through alot of nvrammin here:

 

[Oracle]

Shall I try to flash an older version of the firmware (pre-384) and check this bug?
If yes, can I save my current settings and scripts, so I could flash back to 386 and restore the configuration after I'm done testing?

 

[chongnt]

I just made a proof of concept that shows how a "nvram get" command will hang my script on my AC86U... I didn't get beyond 2245 on my counter (which is literally a minute or so after running this):

View attachment 41826
View attachment 41827

And as predicted... the stuck nvram get is visible in htop... when I kill it, the script continues to run. And now that I'm looking at this, I'm wondering if all these other "nvram get" commands sitting here are also items that have gotten stuck, and never cleared out. <sigh>

#!/bin/sh

i=0

while true; do

    i=$(( $i + 1 ))

    state1=$(nvram get vpn_client1_state)
    state2=$(nvram get vpn_client2_state)
    state3=$(nvram get vpn_client3_state)
    state4=$(nvram get vpn_client4_state)
    state5=$(nvram get vpn_client5_state)

    clear
    echo $state1 $state2 $state3 $state4 $state5
    echo $i

done

exit 0

@dave14305 @SomeWhereOverTheRainBow - thought you'd be interested as well.

Running this version, I got to over 12k and it still running. I noticed both CPU core are 100% and at times the counter slow down a bit but it never get stuck.

2 0 2 0 0
12118
2 0 2 0 0
12119
2 0 2 0 0
12120
^C

 

[Oracle]

Running this version, I got to over 12k and it still running. I noticed both CPU core are 100% and at times the counter slow down a bit but it never get stuck.

Your router model and firmware version?

 

[chongnt]

Your router model and firmware version?

RT-AC86U running 386.5_2
From the sticker on the box, it is Made in China, MFG Year: 2020

 

[Oracle]

Mine was made in 2021.
chongnt, do you have anything from the addon packages?
Any VPNs active?

 

[chongnt]

Mine was made in 2021.
chongnt, do you have anything from the addon packages?
Any VPNs active?

I have quite a number of addons, skynet, YazFi, x3mRouting, Unbound, connmon, ntpmerlin, scmerlin, uiscribe, vnstat, wireguard manager, adguard home. I have two OpenVPN clients and wireguard vpn clients running.

 

[Oracle]

I tried to disable almost everything to check if there was some kind of conflict.
Stopped VPN and Samba, commented out every line from post-mount and services-start, unplugged everything from USB, rebooted several times.

It doesn't matter.
The test script always fails for me. Sometimes as early as 17 iterations, sometimes goes to more than 3000. But eventually hangs.

I tried to plug my SSD drive back in and enable Samba and that just won't work anymore. As soon as I click Apply, the router begins to spawn zombie smbd processes which take 100% CPU. Starts with 1 that takes 100% of 1 core, then 2, now 100% of both cores and then 3 and 4 and so on...

I'm confused how did I manage to break it so badly.

 

[DarkKnight75]

I returned 3 AX86U's after several issues with diff firmware and kept my AC86U. My 2 yr old AC86U was fried in a storm 2 weeks ago and just replaced with another 1. No issues, reliable

 

[Oracle]

@DarkKnight75 , would you try the test script from the first page of this thread?
The one with the 10,000 nvram reads.

 

[Viktor Jaep]

I'm confused how did I manage to break it so badly.

Yeah, running a bunch of nvram lookups shouldn't break anything. You might need to just wipe and refresh, sorry to say. Not sure what may have gone wrong adding/removing tools or settings.

I really appreciate your time (as well as @SomeWhereOverTheRainBow's), but truthfully, the only thing we can do now is just put in a bug report with Asus... which, I don't know would even work, as we're using alternative firmware from the one that they supply these routers with. I bounced this off RMerlin, and he feels we're dealing with a racing deadlock condition, but since it's such an insignificant issue that really doesn't affect its overall stability, there is little chance it would ever be looked at by the Asus devs.

So at this point, let this just be a known issue that needs to be planned for and taken into consideration on AC86U's (maybe not all of them depending on the year/slight hardware differences) in order to maintain the best stability possible for custom scripts.

 

[chongnt]

@Viktor Jaep and @Oracle, perhaps can contact your local Asus support if this qualify for a replacement unit?

 

[Viktor Jaep]

@Viktor Jaep and @Oracle, perhaps can contact your local Asus support if this qualify for a replacement unit?

I really doubt it with it being 2yrs old... I would be leaning to getting an AX86U instead if I had to go through all that trouble. LOL.

 

[Viktor Jaep]

I had to find the PID and kill -9 the process.

Yeah, that was some persistent little loop you created there. Even after it locked I had to use a SIGKILL to terminate it completely.

 

[chongnt]

I really doubt it with it being 2yrs old... I would be leaning to getting an AX86U instead if I had to go through all that trouble. LOL.

I have a very responsive Asus local support. When my RT-N66U failed and wanted to release my Asus ddns account that tied to it, they get it done in a couple of days. You can always sell off your replacement unit.

 

[SomeWhereOverTheRainBow]

I tried to disable almost everything to check if there was some kind of conflict.
Stopped VPN and Samba, commented out every line from post-mount and services-start, unplugged everything from USB, rebooted several times.

It doesn't matter.
The test script always fails for me. Sometimes as early as 17 iterations, sometimes goes to more than 3000. But eventually hangs.

I tried to plug my SSD drive back in and enable Samba and that just won't work anymore. As soon as I click Apply, the router begins to spawn zombie smbd processes which take 100% CPU. Starts with 1 that takes 100% of 1 core, then 2, now 100% of both cores and then 3 and 4 and so on...

I'm confused how did I manage to break it so badly.

Did you try a complete factory reset and reconfigure without restoring from old settings?

 

[Oracle]

I don't expect to be able to get a replacement.

In any case, I think it's reasonable for Asus to investigate this. It's in their own interest.
Maybe it's a batch of faulty nvram chips. They could turn to their supplier for this and take measures not to allow it again.
Or it could be bad CPU design. What if there's a flaw in the nvram read instructions?
Or could be a flaw in their implementation - power supply, EM interference, etc.
Given what they've done with the thermal design* of this router, this may seem minor. Yet it's not so negligible, because if the root cause is not understood, problem may be replicated again.

*This is a big red flag. A company that releases such products worries me (unless it's intentional to shorten the products' life and make the users buy more frequently, in which case it's not incompetence but policy).

@SomeWhereOverTheRainBow: no, I did not do a factory reset.

 

[heysoundude]

Decided I'd setup a sort of online diary about my user experience with the Asus RT-AC86U, a.k.a. Wireless-AC2900.

It has dual-core 1.8 GHz CPU, 512 MB RAM and for what I can tell, good throughput and good WiFi radios.

(If looking for a router in 2022, turns out there are better options for a better price. For example, the RT-AX68U appears to beat it, based on what I read in this forum.)

The way I see it, the AX68 is an AC86 upgraded with WiFi6/802.11ax radios
when I looked at my preferred local retailers last week, a pair of them (which will afford wireless coverage to all but the largest of homes or those with impermeable to radio signals construction materials) cost less than an AX86, the current "big daddy".
if you're meshing or needing multiple APs for coverage, it's totally the way I would go.

 

[SomeWhereOverTheRainBow]

Yeah, that was some persistent little loop you created there. Even after it locked I had to use a SIGKILL to terminate it completely.

Okay when you are locked out what happens when you send the kill signal for the process to "resume"?