Hey all,

I performed a hardware upgrade on one of my servers this morning, and all is well except for that it is now showing an Unraid connect error: NETWORK: queryA ECONNREFUSED mothership.unraid.net CLOUD: Socket closed

This is a 7.2.2 box using the onboard motherboard NIC, which is the same thing I had before (different motherboard but no secondary NIC/wifi). I took the flash backup from Connect before the hardware change, but since I have started the machine, it has not been able to reconnect to the Connect plugin. Network access itself appears fine, as I can reach the docker containers and webGUI.

I changed the DNS settings to the recommended 208.67.222.222 / 208.67.2220.220 and rebooted the machine, but still no change. I saw it recommended in another thread to try this plugin install command via the CLI, this appears to have taken but I still see the same error.

I've attached Diagnostics, so please let me know if there are any recommendations for next steps.

Thanks,

Outlaws

unii-diagnostics-20260110-1221.zip

The diags show that Unraid has connectivity issues, and the fact that you cannot ping 1.1.1.1 confirms there's an external network issue.

Hey Jorge!! I hope you have been well, thank you for taking the time to review this concern. Unfortunately the network on the box does seem functional outside of Connect/the apps tab for a few reasons, if it were experiencing a complete network outage that would be much easier for me to continue troubleshooting.

I did find that the GluetenVPN docker does not function, and doesn't seem to get network access (pings fail as well), but that is the only other hiccup I've seen aside from Connect.

Working on the machine, I can:

1. Access webGUI/SSH/perform file transfers from another LAN host

2. I can access the Plex docker from LAN and WAN

3. The Pihole docker is online, reachable, taking DNS traffic. This docker is NOT configured as the host machines DNS, I know that is not permitted.

4. Yesterday, I was able to access the webGUI and manage the server via https://connect.myunraid.net -> Server Details -> Manage, however this now appears to now be getting stuck on 'Loading'. That is definitely the behavior I would expect, and I found it odd that I was able to access the device remotely yesterday from connect.myunraid.net, where it would still the Connect 'connection error' alert. Apologies that I did not get a screenshot of this behavior.

5. I believe that previous screenshot shows the unraid-api as "status: online" and able to install a plugin over the network, but I may not be interpreting this output correctly.

6. For the ping commands, I believe that output is indicating that ping fails because Connect itself it reporting it does not see a network? I may be misinterpreting that as well, but I can query the box from other network nodes and see ping responses without issue.

Edited January 11Jan 11 by Outlaws

1.1.1.1 is the Cloudfare DNS, if you cannot ping that, either something on your network is blocking it, or there's a problem with your ISP.

Hi Jorge,

I think that I may have found the issue. I noticed that no Docker containers in bridge mode were reachable, only those with an assigned IP in my lan subnet (Plex, PiHole). When I check the network settings, it looks like I'm missing the default route to the gateway somehow? I think that certainly explains the behavior being seen. This 'working' node is still on 7.0.3, so the page looks just a bit different.

I was not able to find a page for this setting in the Docs. Can you confirm that I should enter route:"default", address:"192.168.1.1 via eth0", and metric:"0" in those fields? I can then see if things come back, or let me know if there is another way to correct that.

This node:

Working unit:

Edited January 11Jan 11 by Outlaws

Very cool, that was the problem. I re-added the default route as 'Route 0.0.0.0/0, <gateway IP>, metric 0'. Connect is back online and the dockers have wider network access again.