RT-AC88U crash after 386.9 upgrade from 386.7_2

[Pyjnet]

Hello,

First of all, thank you for this alternate firmware !

Unfortunately I get a router crash almost 10 hours after the upgrade, I have to say i did it several times before posting there and also tried several things without success, so i can also reproduce the issue.

My conf : I had a RT-AC88U as central router/aimesh core and a RT-AC68U for an aimesh node through a gigabit copper link... almost 35 wifi devices, several domotics devices, lights, plugs... amazon devices and 11 cabling devices, i have also a gigabit TP-LINK swich connected on RT-AC88U... was a D-Link, i replaced it recently.

I also use a PI-Hole for DNS running on a RaspBerry and Observium on another RaspBerry to monitor main devices... i have also setup a syslog on internet (SolarWind) where i collect all RT-AC88U log (debug,all).

No script in JFFS.

All that run perfectly in 386.7_2...

I performed a hard reset, a JFFS format and a new manuel setup based on screenshot... 386.9 crash and i don't find in syslog something explaining that.

The only thing i did not do is to start again without my 2 HDD plugged in USB port.

I can try it again without HDDs but need some times again and a WAF compliant time slot because it is my only system an production one

I can share anything needed to help with investigation (log, conf...) if this issue can be considered as a bug and not only an isolated issue.

Please let me know for any questions or more.

Best regards.

Pierre-Yves.

 

[L&LD]

Welcome to the forums @Pyjnet.

There are many things to test for with your hardware and configuration.

First, see the replies to the link below.

AC68 - Hardest Crash Ever

Been running Merlin for a long while. I dont always update to each and every Merlin version, but maybe twice a year. Recently updated from 386_7_2 to 386_9. Dirty update without a router clear and reset and seemed to run fine for a few days. One morning in the middle of talking with Alexa...

www.snbforums.com

Secondly, you may try powering off the router (I would do the entire network) with the power plug pulled out of the router and the power switch left 'on') for a few minutes, and then power it up again.

If the issues continue, do the above again, but before you power up the network, remove the HDDs and test again.

Issues continue? Perform a full reset without the HDDs introduced at all. And also, make sure you only use LAN Port 1-4 on the router (if you aren't doing so already), Using Ports 5-8 is a known (possible) cause of random issues in many Asus 8 port models (and particularly the now 'old' RT-AC88U).

Double check your steps (order is important) in the links below.

Fully Reset / Best Practice Setup / More

 

[Wanabo]

I had the same troubles with my RT-AC88U. After the firmware to upgrade 386.9. I did a restore to factory settings and configured everything new. After a day or two the network was down all of a sudden. Nothing I did could fix the issue. So I took my mesh slave from the attic and reset it to work as a replacement of the RT-AC88U. At least now I could use the network and internet again. After a day (power adapter was unplugged) I reconnected the RT-AC88U and to my surprise and re leave it functioned 100%.

So the hint from L&LD powering of your device for a longer period is what I also did and helped me.

 

[Tech9]

After a day or two the network was down all of a sudden.

Power down won't help. This firmware has some intermitted Dnsmasq issue from what I can tell. I don't know why and when it stops working, but I had to restore one AC86U, another user had locked AC86U, yours is second AC88U report for the same. I don't have the hardware to investigate further.

 

[Pyjnet]

Welcome to the forums @Pyjnet.

There are many things to test for with your hardware and configuration.

Hello L&D,

Thank you for your reply.
Some of the "things" but not all have been done so i will try again following strictly your advises... i will have a timeslot next Friday to do the job and let know about the result.

Like i already wrote, keep in mind that i used each version/release of Merlin before that one and all run perfectly on this hardware except this one and a "dirty" rollback to 386.7_2 solves !

I forgot to share the status when the crash had occurred, already running wired appliances are still connected and can reach internet normally (Apple TV, my laptop...), wifi devices are all out of services despite the fact SSID is still flooded but impossible to connect.

RT-AC88U is unreachable by HTTP or HTTPS (SSH is disabled in conf), IP and port are discovered by "angry scanner", a hard reboot solves waiting the next crash and is the only way to recover.
It could be that some threshold was modified or that some counter for any reason reaches a limit, thing which noit happen in other Merlin versions/release.

Thank you to Wanabo to share you experience ! did you loose both IP connection though Ethernet and wifi ?

Power down won't help. This firmware has some intermitted Dnsmasq issue from what I can tell. I don't know why and when it stops working, but I had to restore one AC86U, another user had locked AC86U, yours is second AC88U report for the same. I don't have the hardware to investigate further.

Thank you Tech9 "This firmware has some intermitted Dnsmasq issue", very interesting to investigate.

 

[Pyjnet]

Hello,

I did the job this morning (in France...) so i did the total reset procedure directly without USB HDD and with a long global power off also and setup manually again with 386.9.0 but did not plug again the USB... i will definitively handle them with another appliance...
For now all is fine, i have to wait if the crash occurs again or not, i'll let you know... if all is running tomorrow, it would be a good signal.
Syslog on papertrailapp.com is started with "debug,all", so let's go.

 

[Pyjnet]

I plugged again after a save of .cfg the USB HDDs at noon French time so now almost 6 hours ago to test that... just to wait at least tomorrow, all is running fine for now.

 

[Tech9]

This 386.9 firmware is the first "honest" one with straight "out of memory" record in logs before it crashes. Your HDD is perhaps causing your issues.

Release - Asuswrt-Merlin 386.9 is now available for AC models

I think you've made your thoughts on ASUS/Merlin quite clear here already - https://www.snbforums.com/threads/gt-ax6000-vs-rt-ax86u.78498/page-3#post-816584 I do wish people would stop reacting/responding to these posts If he actually has filed these reports and Asus is not responding, I see...

www.snbforums.com

Release - Asuswrt-Merlin 386.9 is now available for AC models

I've had this issues for a bit, any version after 386.5_2 forces on trend micro's AI protection even with it off, on firefox it causes a lot of SSL_ERROR_RX_RECORD_TOO_LONG errors when you got on sites trend micro has decided are malicious. What do I do to fix it? I'm using an RT-AC86U. I just...

www.snbforums.com

My advice - use Asuswrt-Merlin 386.7_2 instead on your router. I had to restore one AC86U about a week after 386.9 upgrade.

 

[Pyjnet]

I think also that is the root cause... I intend to use an external homemade NAS and/or to switch on newer Asus router... unfortunately i did not have this "out of memory" in logs before crash but may be i missed it, right now, all is tracked by Observium (snmp poller) and also an eternal syslog so normally i should get some warning in case of new crash.
I also tuned the configuration to suppress any feature i don't need anymore, for now all is fine, we'll see. thank you very much to share

 

[Pyjnet]

 

[naznaznaz]

Hi, go back to 386.7_2 and do a clean install. For clean install; After installing 386.7_2, turn the device off and on again for 10 - 15 seconds using the on/off button, then reset to factory settings. When performing a factory reset, also tick the box next to the "factory reset" button. After the reset, re-enter all your information and perform a clean installation.

 

[Pyjnet]

Hello, I share the RAM consumption increase after noon on my RT-AC88U with 2 USB HDD plugged and setup, before noon, no USB HDD plugged, i didn't get (yet) an "out of memory" error in logs...we'll see.

 

[Pyjnet]

Hello, did you intend to post there ? because it is not at all the step of this topic...

Hi, go back to 386.7_2 and do a clean install. For clean install; After installing 386.7_2, turn the device off and on again for 10 - 15 seconds using the on/off button, then reset to factory settings. When performing a factory reset, also tick the box next to the "factory reset" button. After the reset, re-enter all your information and perform a clean installation.
View attachment 47917

 

[naznaznaz]

I wrote the healthiest way to return to your last working state.

 

[Pyjnet]

Thank you, but if you read the full post, you'll see i has been already done several times now i just share the final results in case it could help someone else or to understand.

 

[Pyjnet]

It remains quite correct... thanks to everyone for your help, i hope also it will be helpful for someone else