Install watchcat on the router and configure it to "Run Script". Set some short monitoring interval, "host to check", etc.
Create your own simple script that only makes a log entry:
#!/bin/sh
logger -t watchcat "No response received"
Once the router detects a connection problem, you should get a log entry.
Next time you encounter a problem on your PC, check your router log for the same time period.
just now I think the issue happened again (I get server not found in firefox when it happens,I think )
so I checked openwrt log and I saw this (replaced the mac addresses):
Sun Oct 19 15:38:20 2025 user.notice nft-qos-monitor: ACTION=remove, MACADDR=odroidmacaddr, IPADDR=192.168.1.102, HOSTNAME=odroid Sun Oct 19 15:38:21 2025 user.notice nft-qos-dynamic: ACTION=remove, MACADDR=odroidmacaddr, IPADDR=192.168.1.102, HOSTNAME=odroid Sun Oct 19 15:38:24 2025 user.notice nft-qos-monitor: ACTION=remove, MACADDR=laptopmacaddr, IPADDR=192.168.1.101, HOSTNAME=laptop Sun Oct 19 15:38:25 2025 user.notice nft-qos-dynamic: ACTION=remove, MACADDR=laptopmacaddr, IPADDR=192.168.1.101, HOSTNAME=laptop
the odroid device is connected via lan (the only lan execpt wan port) and laptop is connected via wifi. there are other devices on wifi connected to router but there are not in the log that is printed at the time of bug.
the other think is that QoS over Nftables is not enabled on my device and this log is the only thing that openwrt prints out when the issue happens (I think, because they I saw it twice now)
I will remove QoS over Nftables and check to see if I get the error again or was that log unrelated or related but not the cause just the symptom.
Then you need to make sure FF is configured to use regular "system" DNS and your PC is using your OpenWrt router as a DNS server.
In order to see what's going on with the name resolution in OpenWrt you can enable logging in "DHCP and DNS" section in Luci.
I meant I cant check the issue to see it happen because it is random. but one the sysmptoms of the internet not working is that firefox giving that error which then I can go check openwrt log to see the log.
the issue is not relevant to firefox.
I get ping timeout to 8.8.8.8 and any other (even national IPs).
Then do the same through watchcat on the router as suggested earlier.
If you observe loss of upstream connectivity at regular intervals, it is likely due to periodic reconnection between your LTE modem and your carrier.
the isp own modem has web interface with ping ability.
I ran it from there and it has the timeouts there randomly too.
can you guys tell me if this is the case:
and watchcat:
Sun Oct 19 17:58:24 2025 daemon.info watchcat[6195]: Could not reach 8.8.8.8 via "wan" for "1" seconds. Running script after reaching "1" seconds Sun Oct 19 17:58:24 2025 daemon.info watchcat[6195]: Running script "/etc/watchcat.user.sh" for network interface: "wan". Sun Oct 19 17:58:24 2025 user.warn watchcat: cfg01e014: pingperiod cannot be greater than 1. Defaulted to 0 seconds (1/5 of period) Sun Oct 19 17:58:34 2025 daemon.info watchcat[6195]: Could not reach 8.8.8.8 via "wan" for "10" seconds. Running script after reaching "1" seconds Sun Oct 19 17:58:34 2025 daemon.info watchcat[6195]: Running script "/etc/watchcat.user.sh" for network interface: "wan". Sun Oct 19 17:58:34 2025 user.warn watchcat: cfg01e014: pingperiod cannot be greater than 1. Defaulted to 0 seconds (1/5 of period) Sun Oct 19 17:58:49 2025 daemon.info watchcat[6195]: Could not reach 8.8.8.8 via "wan" for "1" seconds. Running script after reaching "1" seconds Sun Oct 19 17:58:49 2025 daemon.info watchcat[6195]: Running script "/etc/watchcat.user.sh" for network interface: "wan". Sun Oct 19 17:58:49 2025 user.warn watchcat: cfg01e014: pingperiod cannot be greater than 1. Defaulted to 0 seconds (1/5 of period)
the timeout just comes and then goes away.
and this is the ping part on modem that I think it happens:
64 bytes from 8.8.8.8: seq=140 ttl=110 time=58.460 ms 64 bytes from 8.8.8.8: seq=141 ttl=110 time=109.401 ms 64 bytes from 8.8.8.8: seq=142 ttl=110 time=733.870 ms 64 bytes from 8.8.8.8: seq=143 ttl=110 time=19252.032 ms 64 bytes from 8.8.8.8: seq=144 ttl=110 time=18668.973 ms 64 bytes from 8.8.8.8: seq=145 ttl=110 time=18262.820 ms 64 bytes from 8.8.8.8: seq=146 ttl=110 time=17673.607 ms 64 bytes from 8.8.8.8: seq=147 ttl=110 time=16971.372 ms 64 bytes from 8.8.8.8: seq=148 ttl=110 time=16389.218 ms 64 bytes from 8.8.8.8: seq=149 ttl=110 time=15591.600 ms 64 bytes from 8.8.8.8: seq=150 ttl=110 time=14725.292 ms 64 bytes from 8.8.8.8: seq=162 ttl=110 time=2800.300 ms 64 bytes from 8.8.8.8: seq=163 ttl=110 time=1811.312 ms 64 bytes from 8.8.8.8: seq=164 ttl=110 time=1292.063 ms 64 bytes from 8.8.8.8: seq=165 ttl=110 time=605.081 ms 64 bytes from 8.8.8.8: seq=166 ttl=110 time=1115.588 ms 64 bytes from 8.8.8.8: seq=167 ttl=110 time=1089.109 ms 64 bytes from 8.8.8.8: seq=168 ttl=110 time=1095.096 ms 64 bytes from 8.8.8.8: seq=169 ttl=110 time=958.735 ms 64 bytes from 8.8.8.8: seq=170 ttl=110 time=754.846 ms 64 bytes from 8.8.8.8: seq=171 ttl=110 time=959.357 ms 64 bytes from 8.8.8.8: seq=172 ttl=110 time=1032.688 ms 64 bytes from 8.8.8.8: seq=175 ttl=110 time=1093.014 ms 64 bytes from 8.8.8.8: seq=176 ttl=110 time=923.104 ms 64 bytes from 8.8.8.8: seq=177 ttl=110 time=240.576 ms
257 packets transmitted, 242 packets received, 5% packet loss round-trip min/avg/max = 42.240/803.847/19252.032 ms
so 15 packets gone and 12 of them are one behind another.
Am I correct to think that it is my internet and not my own devices?
another question is that can my own download upload choke the modem like this? my opinion is that it cant do it because modem itself the source of my inernet.
Sun Oct 19 18:06:16 2025 daemon.info watchcat[6195]: Could not reach 8.8.8.8 via "wan" for "3" seconds. Running script after reaching "1" seconds Sun Oct 19 18:06:16 2025 daemon.info watchcat[6195]: Running script "/etc/watchcat.user.sh" for network interface: "wan". Sun Oct 19 18:06:16 2025 user.warn watchcat: cfg01e014: pingperiod cannot be greater than 1. Defaulted to 0 seconds (1/5 of period) Sun Oct 19 18:06:26 2025 daemon.info watchcat[6195]: Could not reach 8.8.8.8 via "wan" for "10" seconds. Running script after reaching "1" seconds Sun Oct 19 18:06:26 2025 daemon.info watchcat[6195]: Running script "/etc/watchcat.user.sh" for network interface: "wan". Sun Oct 19 18:06:26 2025 user.warn watchcat: cfg01e014: pingperiod cannot be greater than 1. Defaulted to 0 seconds (1/5 of period)
@reza You could double check the network cables from the wall outlet to your modem and from your modem to your router. Just unplug them and put them back in. In case you have spare cables try to exchange them and see if that helps.
the modem is power via POE (the same ehternet cable that brings data from modem to openwrt router).
the modem is not restarted. so I don’t think it is power related.
I have manually restart modem many time to see if it fixes it (from its web interface not cable removal)
so I don’t think it is hardware issue.
I think it is internet (ISP interanet) . Internet in Iran is getting the fiber upgrade but it has a long way to go and ISPs are shitty as f…k and very unresponsive. thought if internet users are to be believed shitty ISP behavior is universal (or at least shared between iran and usa).