Internet intermittent off for hours but Tailscale subnet router on Brume still connected

Routers Cellular, USB modems
1

I have two sites (home and farm) connected using Tailscale. Each site has a couple of Win 10/11 PC’s and several devices that cannot use Tailscale so I have a Brume at each site operating as a Subnet Router.

Home has residential FTTP, farm has 5G cellular on a TP-Link NE210 Outdoor router. The 5G router is set in a Bridge mode and has local ip 192.168.254.1. If it is relevant I am in the UK.

It’s a simple setup, Home uses 192.168.3.0/24 for the Tailnet and Farm uses 192.168.50.0/24. The Blume is the gateway at each site. The devices on each site can all communicate over Tailscale VPN without any problem or issues.

The problem is at the Farm site the devices loose connectivity to the Internet at least once per day for a continuous period of several hours. It only affects PC’s which as a result have no internet access even if I turn on the Tailscale app on the PC (I usually don’t have it turned on and just use the Subnet Router. The strange thing is the Tailscale VPN between Home and Farm stays connected and Home and Farm devices can access each other so the Cellular Internet service is up. Internet connectivity will return eventually or if I restart the Farm Blume it will return. If I check on he Tailscale web console the Farm Blume shows Not Connected even though the Tailscale VPN is up,amd working

If I connect to one of the Win PC’s at Farm and try to browse to a website I get Webpage Not Found / DNS error. Command line ping to a known site like www.cisco.com returns host not found and ping 8.8.8.8 returns no response. Normally both test would work without problem.

Browsing from a Farm PC to the 5G router web interface on 192.168.254.1 which is connected via the Brume WAN interface also fails. This would normally be accessible if there was internet access from the Farm LAN.

I am thinking this is a routing problem on the Brume but no idea of the cause. Or it might be. Tailscale problem, I am a bit stuck. Any help or ideas please?

Also the problem only affects Farm and not Home. Blume devices were purchased this week and running latest version.

Thanks in advance.

3

Update if it provides any clues, to day the no internet from Farm LAN fault occured so I SSH’d into the Farm Brume and ran “ip route” then noted the default route for the Brume eth0 (WAN) is via the TP-Link /30 bridge subnet.

I restarted the Brume and internet connectivity returned, SSH’d into the Brume and ran “ip route”. The default route now on eth0 was via 192.168.254.1 which is the TP-Link 5G router LAN subnet.

The Brume WAN interface (eth0) is configured to use DHCP getting an address from the TP-Link.

I assume the Brume is learning the default route via the DHCP address assigned by the TP-Link but no idea why is changing between the TP-Link Bridge subnet and TP-Link LAN subnet. If I had left it I expect it would have fixed itself after several house.

I am thinking change the TP-Link to Router mode but technically this seems like a wrong choice as it should be “bridge mode” but maybe I am missing some finer detail.

Any ideas?

4

Hi

You might try disabling the TP-Link's DHCP in router mode, then switching to bridge mode to see if any issues persist.

If problems remain, you may need to configure the TP-Link in router mode to create a DMZ for Brume.
This should achieve a similar function to bridge mode.

1 Like