Flint 2: drops.. overnight?

Flint 2, latest firmware, the router is connected to an ONT from my ISP. Everything works perfectly during the day but when I wake up, there's no internet connection. As in, no device can access the internet despite the status in the router app showing green, connected. A third day in a row. Rebooting the router (I even scheduled a reboot) does not fix the problem.

I accidentally discovered that changing any DNS setting and then back to auto fixes the problem. So I suppose it's related to the dynamic IP lease? Sorry, that's the troubleshooting I was able to do. Anything worth changing to see if it helps?

  1. May I know the Flint2 WAN protocol to ISP ONT is DHCP or PPPoE?
  2. Please SSH to the router when the issue reproduced, execute these commands to check, and post the result:
ifconfig
ip r
ping 8.8.8.8
ping www.google.com
ethtool eth1
  1. Please SSH to the router when the issue reproduced, manually restart the WAN port to see if it restores to work:
    ifup wan

  2. Please PM me the issue syslog.

Hey Bruce, it's DHCP.

In the classic style of Murphy's law, this morning the issue didn't occur (for the first time in 5 days) :slight_smile:
I'll see if it's the kind of issue that solves itself.

1 Like

Hi, I have the same exact issue as OP.

I wake up in the morning, and all of the internet traffic in my Flint 2 is not working and loads nothing. I log into the router's web interface (either with desktop or iOS Gli.net app), and it says it's "online" with green status. I fiddle with the DNS settings (I either swap around the DNS addresses or do nothing with the form), press "submit", and suddenly everything works...?

I had a feeling it was related to DHCP, but what do we do with DHCP to make sure the router stops doing this every morning?

It probably doesn't relate to DHCP.
If DHCP does not work, you may not be able to login directly to the GL GUI or the gllinet APP without setup the static IP.

When it reproduces, please check it out according to these:

Also having the same issue here, always in the morning but not every day.

Flint2 connected to a 5G mobile internet modem, DHCP.

When this happens, ping 8.8.8.8 works on the router but I have no connectivity on my PC (ethernet) or wifi.

ifup wan restores connectivity.

Hmm, it is possible the isp dns is down.

Then you won’t be able to resolve domain names, this is quite common with network issues, you could set a different dns as long pinging to ips work, if not it will fail, then the isp is down likely.

or the main connection line has issues, like with fiber a unclean fiber optic could be a issue, but then you often notice more issues and unstable internet.

if the issue is only dns, you may not want to use DoT or DoH first since often these services use a domain name first, and I don’t know how it uses it in those gl scripts, first you want a working dns and then apply DoH or DoT, just in case, it might just work but I haven’t checked it.

It could be IPV6, imo it has alot of adoption problems, could be also on the isp side.

It would be something on your own network, if full dhcp loses, then you may want to investigate for turning down dhcp rebind protection if there is a second router, a ip conflict, or a weak PoE accesspoint/switch reaching its max.

I had two situations with unifi switches, sometimes they can claim a low segment ip when the AP is there, it can give very unexpected and slow network results, since a switch is also layer 2 you sometimes don’t necessary see the issue by forehand funny enough you can even receive vlans and not being able to access the switch ui (appearing offline), having a sudden disconnect and then working again.

And I had a maxed PoE connection, it was to low power left, but enough to keep my ap running with no uptime lost, only the wifi stops working when speedtesting.

These are things to look out for if the more normal things above are not happening.

1 Like

It happens whether I use the ip DNS or use a custom DNS (cloudflare).

My setup is pretty basic, just 5G Modem → Flint 2 and that’s it.

Can the issue be isp side when ifup wan or a router reboot always restores connectivity?

This is possible, but I would think if it doesn’t renew from itself it might be a bug in the version of OpenWrt inside the gl firmware.

I know these issues did exist, and sometimes also re-appear on OpenWrt, the only way to be exactly sure it is external is understanding the protocol from your isp.

Is it dhcp?, look for a log message containing NAK from dnsmasq, then the isp was rejecting dhcp requests, often associated with isp issues.

With pppoe that is often a padding error, or the server cannot be found.

It is also possible it is some rate limiting policy, and after some time it disconnects you.

None of you guys ever mentioned their router firmware version.

1 Like