Same error. 2026-08-26T18:43:54-05:00 ERROR default route not found: in 4 route(s) 2026-08

Posted by JoeFCards

My question: Same error. 2026-08-26T18:43:54-05:00 ERROR default route not found: in 4 route(s) 2026-08-26T18:43:54-05:00 INFO Shutdown successful SparkBox AI replied: The credentials didn't fix it — default route not found is a network-level problem, not auth. Let's check if the Docker network itself is broken. Step 1: Check if the VPN network exists bash docker network ls | grep sbproxy If that returns nothing, the network is gone. That's the problem. Step 2: Check the container's network mode bash docker inspect sb-gluetun --format='{{.HostConfig.NetworkMode}}' Should say sbproxy or container:sb-gluetun. If it says default or bridge, that's wrong. Step 3: Check your host's routing table bash ip route | head -10 Look for a line starting with default via. If it's missing or shows multiple defaults, that's the issue. Step 4: The nuclear fix (if the network is wedged) bash sudo sparkbox down sudo sparkbox up This tears down and rebuilds all Docker networks. Safe — your data stays put. Run those first three commands and paste the output. If docker network ls shows no sbproxy, skip straight to step 4. What I still need help with: Everything seems to check out, but nothing has changed. Glutun and vpn connection still wont work.

3 replies

Chris wrote:

This has gone on too long, and I'm sorry. Here's what's certain: your Surfshark key and address are correct — this error isn't about them. It means Docker on Windows lost the network route it needs, so nothing you changed in SparkBox could fix it. The missing step is to fully restart Docker: right-click the whale icon in the system tray, choose Quit Docker Desktop, reopen it and wait until it says running, then run sudo sparkbox restart media. If it still won't connect, that's a Windows-side limit Tom has flagged — a Linux box or NAS runs it reliably.

JoeFCards wrote:

I used the windows .exe to install this though not docker.

Chris wrote:

Right — the .exe runs SparkBox inside Windows' built-in Linux layer, not Docker Desktop, so the earlier tray step doesn't apply to you. That error means that Linux layer lost its network route, and nothing in SparkBox can repair it. One step does: open PowerShell on Windows, run wsl --shutdown, wait about ten seconds, then reopen SparkBox. That restarts the Linux layer cleanly and usually brings the VPN straight back. If it still won't connect, I'm flagging this thread to Tom — two weeks on this is too long, and a Linux box or NAS is where SparkBox runs most reliably.