Network down suddenly > power reset only

[Joern]

Hello,

my openWRT seem to get stuck after sometimes 12h or 2 days, but cannot find easily a root cause. Whole network is then down and even connecting by LAN to openWRT, I get no ip.

GL.iNet GL-MT6000
OpenWrt 23.05.5 r24106-10cc5fcd00 / LuCI openwrt-23.05 branch git-24.264.56413-c7a3562

> Only power on/off helps, everything boots as normal.

1. can I check/activate further/other log to better understand?

I installed watchcat to monitor with ping to internet, to restart in case, but have not seen a restart due to that. After crash this morning, I only found in log:

Thu Feb  6 08:03:34 2025 daemon.warn odhcpd[1815]: No default route present, overriding ra_lifetime!
Thu Feb  6 08:03:35 2025 daemon.notice netifd: Interface 'Wireguard' is setting up now
Thu Feb  6 08:03:35 2025 daemon.notice netifd: Interface 'wan6' is now up
Thu Feb  6 08:03:35 2025 daemon.notice netifd: Interface 'Wireguard' is now up
Thu Feb  6 08:03:35 2025 daemon.notice netifd: Network device 'Wireguard' link is up
Thu Feb  6 08:03:35 2025 user.notice firewall: Reloading firewall due to ifup of wan6 (eth1)
Thu Feb  6 08:03:35 2025 user.notice firewall: Reloading firewall due to ifup of Wireguard (Wireguard)
Thu Feb  6 08:03:35 2025 user.notice ddns-scripts[3792]: spdyn: PID '3792' started at 2025-02-06 08:03
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: reading /tmp/resolv.conf.d/resolv.conf.auto
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using nameserver 185.22.44.50#53
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using nameserver 185.22.45.50#53
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using nameserver 2a00:aaaa:bb00::1#53
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using nameserver 2a00:aaaa:bb00::1#53
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for test
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for onion
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for localhost
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for local
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for invalid
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for bind
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: using only locally-known addresses for lan
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: read /etc/hosts - 12 names
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: read /tmp/hosts/dhcp.cfg01411c - 7 names
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq[1]: read /tmp/hosts/odhcpd - 12 names
Thu Feb  6 08:03:36 2025 daemon.info dnsmasq-dhcp[1]: read /etc/ethers - 0 addresses

[frollic]

where does WG come into the picture ?

[Joern]

good question, I have wireguard and adguard running but WG was not used at that time as I'm at home. Maybe only my android (WG tunnel) auto-connected when local wireless was lost.

Surprisingly there was not log the 30min before, so when recognized crash at approx 08:05 with laptop loosing WIFI & internet, I assumed data from 08:03 is relevant one

From: Thu Feb 6 08:03:36 - looks like dnsmasq did restart.

[Joern]

Again today after little over 24h a crash.
Using watchcat and reboot after 24h, no issue last days. Today I change to 26h to move forward midnight and not reboot during the day. Then at 24,xx h a crash. Only reboot via shh worked. Openwrt Website was not loading anymore properly:

Log I collected on log server:

2025-02-13T21:27:42+00:00 OpenWrt AdGuardHome[3251]: 2025/02/13 20:27:34.662762 [error] dnsproxy: [::1]:5353: response received over udp: "dialing [::1]:5353 over udp: dial udp [::1]:5353: i/o timeout"
2025-02-13T21:27:44+00:00 OpenWrt AdGuardHome[3251]: 2025/02/13 20:27:32.235670 [error] dnsproxy: 8.8.8.8:53: response received over udp: "dialing 8.8.8.8:53 over udp: dial udp 8.8.8.8:53: i/o timeout"

2025-02-13T21:28:13+00:00 OpenWrt AdGuardHome[3251]: 2025/02/13 20:28:07.075031 [error] dnsproxy: upstream 8.8.8.8:53 failed to exchange ;1471500108.rsc.cdn77.org.#011IN#011 AAAA in 1m50.283150447s: dialing 8.8.8.8:53 over udp: dial udp 8.8.8.8:53: i/o timeout
2025-02-13T21:28:18+00:00 OpenWrt AdGuardHome[3251]: 2025/02/13 20:27:59.621434 [error] dnsproxy: upstream [::1]:5353 failed to exchange ;1471500108.rsc.cdn77.org.#011IN#011 A in 1m48.52585396s: dialing [::1]:5353 over udp: dial udp [::1]:5353: i/o timeout
2025-02-13T21:28:20+00:00 OpenWrt AdGuardHome[3251]: 2025/02/13 20:28:12.195562 [error] dnsproxy: 8.8.8.8:53: response received over udp: "dialing 8.8.8.8:53 over udp: dial udp 8.8.8.8:53: i/o timeout"
2025-02-13T21:28:21+00:00 OpenWrt procd: - shutdown -

[frollic]

I'd start by disabling AGH, it's a memory hog.

[Joern]

ok will then stop service AGH, lets see if runs longer and report

[Joern]

well, AGH disabled on openWRT and 24h are passed now without a issue! I moved meanwhile AGH to a existing proxmox sever which is also a simple solution and running fine there so far...

So will keep it running and maybe try 24.10 soon. Without AGH mit Router is fairly sleeping and RAM is empty, but if it runs like this is ok...

[Joern]

After 4days no issue.

AGH seems to be root cause. Let's see, if it will improve with next Rev.