WebUI not working after a while (RT-AC86u)

[zz63]

I've discovered that if 'Auto Logout' (Administration, System, Basic Config) is set to a large value (960 minutes), the web UI stays logged in, as intended, and httpd does not crash. Setting the Auto Logout to zero (0) will cause httpd to crash after some random amount of time. Try setting Auto Logout to a large number.

I had the automatic shutdown set for 20 minutes. Thanks, I'll try it.

 

[el pescador]

Happens on the AC88U. A lot.

 

[Willing to...]

I don't think that is related to merlin firmware since on my router happened before merlin was installed.
The same was also problem on my ISP provided router so for me it is OS related (various kind of protections probably).
My solution is that I have an old old laptop with WinXP installed and firefox I think version 17 - with them I can log in EVERY time without problem.
If I run symantec endpoint protection I can't log in so I must disable it otherwise when pressing log in the same screen is returned over and over again.
Recently I use portable ungoogled chromium without any extensions to log in and when no one is connected to router mostly I can log in but when some other device is wirelessly connected I mostly can't log in or web ui is broken.

 

[Rando]

I have the same issue on AC-3100 with Merlin 384.19

 

[josh3003]

0 4 * * * service restart_httpd

How would I go about setting this up? I understand it needs to sit within /jffs/scripts/services-start but I don't get the whole cru at the beginning would just 0 4 * * * service restart_httpd work?

 

[cooloutac]

I should add that when the WebUI is unresponsive the router is otherwise still fine (clients can still access the Internet, SSH works etc.).

Have noticed RAM usage unusually high for the past few days (only started using the router 4 days ago and still tweaking it). It's now about 65%. Don't use Trend Micro yet, though was planning to use it for time-based access for some of the clients.

Any suggestions here? Using Diversion & YazFi as well.

Not sure if this the issue in your cause, but this can happen if the router gets stuck while checking for updates. happens on all asus routers I ever tried as long as I can remember. (ac66u_b1, ax58u, ac86u.) Don't know why it was never fixed, maybe its an ISP conflict. no idea. I turn off auto updates and just check the firmware on the asus website and manually upload when needed. checking the trend micros signatures doesn't cause issue that i know of. 65% ram is normal for these routers even if you don't have many features enabled.

 

[JoeBee]

Just wanted to say thanks to this thread and all post lots of useful information. I have had this issue with my rt-ac86u since day 1 and it drove me up the wall frankly, I thought it was something I was doing wrong.

Another more recent October 2022 thread is found here mentioning scMerlin could help with this issue.

Got scMerlin installed now so hopefully I have better luck next time, guess if it does not work I can try to reboot the Openvpn service and even router from putty, at least the option to do something rather then no Asus router webui!

 

[zman888]

I've root caused the issue at least for my case. Here is the thing, there are two sqlite db files hidden inside /jffs/.sys hidden directory and somehow not caught by either dh command or du command no matter where you run the command.

These files were used by traffic analyzer and network monitor. So over time they will fill up your jffs, and cause it to freeze. any operation that involves reading or writing to /jffs will get stuck. This also explains why someone was able to 'kind of' fix the issue by disabling traffic analyzer or a firmware re-flash which will possibly include a jffs wipe. However none of these are permanent solution as the db files will grow back like weed.

So long story short, here is the solution:

Instant fix - run this command when ssh in as admin:
rm -f /jffs/.sys/*/*.db

Permanent fix:
opkg install cron, then depends on the cron version, you could either use
- "crontab -e"
- or directly edit file "/opt/etc/crontab",

add this line: it will delete the 2 db files weekly.
0 0 * * 1 rm -f /jffs/.sys/*/*.db