Constant Log Spam "rc_service: watchdog 247:notify_rc start_cfgsync" on RT-AC68U

[rkk2025]

Hi, this seems to have been going on for a while now, I don't know with which upgrade it started, but it still happens on the lates AsusWRT Merlin Firmware 384.10_2, altough it has been happening already in previous versions. A complete factory reset and reflash didn't help at all.

It seems like every few seconds another line is added in the Log saying "rc_service: watchdog 247:notify_rc start_cfgsync". It also seems to happen at the same time a CPU spike happens. What I can see with "htop" is that CPU usage is rather low 0-1% and then a process called "watchdog" appears putting the CPU at around +-30%, and right after that a process called "mtdblock3" uses around 50% or more CPU for a short amount of time. And this keeps happening again and again every few seconds.

Any idea what could be causing it and how to fix it?

 

[appleseed]

It's an odd problem that hits some users with no clear solution I can see. An Asus tech said it was fixed but it comes back from time to time: https://www.snbforums.com/threads/r...04-notify_rc-start_cfgsync.46274/#post-401757 . So now you know what you're in for.

Could be (low probability) internal flash going bad: (Netgear forum) https://community.netgear.com/t5/Ge...vernight-with-high-SY-and-high-WA/m-p/1504484 , or it could be an Asus/Trend Micro closed-source problem, or it could be AI Mesh related: https://www.snbforums.com/threads/p...builds-are-available.47385/page-4#post-413920

This doesn't answer your question but it also happened as recently as January Alpha build so at least you are not alone, which can be helpful with troubleshooting, or not: https://www.snbforums.com/threads/384-9_alpha-builds-testing-all-variants.54050/page-19#post-461023

A complete factory reset and reflash didn't help at all.

I use Merlin with minimal settings and I don't have this issue so I can only think the best answer for now is to try it again. Unplug all external drives, hit the factory reset, do a minimal config, no VPN, wait... view the logs, wait...wait... view htop/ logs,... and monitor htop/logs again while slowly turning your settings back on (and then plugging in your externals). The error should show itself again, or it won't.

Factory reset appears to work for this member: https://www.snbforums.com/threads/f...to-work-on-dsl-ac68u.48702/page-3#post-463980

I would suggest leaving anything VPN, AI Mesh, or Trend Micro off until last to help narrow your search.

 

[appleseed]

Btw,
is it a DSL or TM AC1900 converted router ?

Also see this from Merlin: https://www.snbforums.com/threads/mtdblock3-cpu-usage-every-10-seconds.38764/#post-320195

If you have Entware setup, you can install "lsof" to help troubleshooting things:

opkg install lsof

Otherwise, take a look at the "top" and "ps" output. Syslog might also contain clues.

Thanks for showing that entware can install lsof, I was able to narrow down the culprit to "lighttpd-monitor" it was constantly using mtd3, and once killed, stopped it.

 

[sugar0]

i have the same problem with an DSL-AC68u!

 

[kernol]

i have the same problem with an DSL-AC68u!

Then it seems to me - you also have the solution in the threads above. Give them a try - especially the factory reset following the M&M guide).
NB - This is not a firmware specific issue - so no point continuing the posts in the DSL-AC68U GNUton thread on this issue.