Hello,

I am experiencing a network connectivity issue when connecting to my Unraid server via WireGuard VPN. While I can successfully access the Unraid server itself (192.168.86.246), I cannot reach any other devices or Docker containers on my local network (e.g., 192.168.86.2:3000 for AdGuard, 192.168.86.3 for NGINX Proxy Manager, or other Docker containers with custom IPs).

This issue started occurring after updating Unraid to version 7.1.2. My WireGuard VPN Manager is configured with "Local server uses NAT: No," and my router has a static route configured to direct traffic for the WireGuard subnet to the Unraid server.

One thing I noticed in VPN manager is that the Remark for static routing doesn't recognize my Unraid address.

I am pretty confident this needs to say Remark: configure your router with a static route of 10.253.0.0/24 to 192.168.86.246.

I tried using AI to troubleshoot this and Gemini seems to think there's some monkey business going on with MASQUERADE, where the NAT setting is not working at all.

Also same issue with NAT on:

Hello,

I had Downgraded down to 7.0.1 and the issue was fixed immediately, but then I recently just upgraded to 7.1.4 and the problem returned. Has anyone experienced this? I can't be the only person.

Hi, glad I found this topic, because I experience this too. And I'm using NAT.

My VPN used to work flawless, but I guess they broke something with UnRAID 7.1.4, because it does not work anymore. I can still do things inside the server, but not on the rest of the network and not on the internet. Just like you. And yes, I use Remote tunneled access and I have Host access to custom networks enabled.

But I guess it's not enough to ask a question here, I think I will have to file a bug.

I filed a report about this bug.

Edited October 13, 2025Oct 13 by RJ59

Hi,

Thank you for filing a report. I was tempted to update to the beta to see if this has been resolved, but I haven't had the time to test things out recently. Currently on 7.0.1 and everything works fine there.

Hi everyone. Just out of curiosity, are you using the Wireguard packaged with Unraid? I ask, because I run a Wireguard instance on OpenWRT and the routing problem for anything (including the webgui) Unraid related is persistent here too. I cannot access from a VPN network (road warrior), via OpenWRT Wireguard any resources on the Unraid host. I can perfectly access my openWRT webGUI on the same network. The problem is really imposed upon me by Unraid. I have tried serveral things in OpenWRT (masquerading, forwarding...), all to no avail. I also tried on the CLI to add my VPN subnet to the iptables for DOCKER-USER, but this solved nothing: iptables -I DOCKER-USER -s 10.10.10.0/24 -j ACCEPT. Being new to Unraid, I was wondering if anyone could explain the mechanisms behind this blocking and where or how we are supposed to influence it? I find going in via iptables a little bit dark and cumbersome (coz I don't understand it haha). Wish ufw were available, or a GUI. Are we not supposed to manage networking at the firewall layer as Unraid users? That could be worrying, coz I like to think I have control ;). Thanks.

15 minutes ago, dmindfuse said:

Hi everyone. Just out of curiosity, are you using the Wireguard packaged with Unraid? I ask, because I run a Wireguard instance on OpenWRT and the routing problem for anything (including the webgui) Unraid related is persistent here too. I cannot access from a VPN network (road warrior), via OpenWRT Wireguard any resources on the Unraid host. I can perfectly access my openWRT webGUI on the same network. The problem is really imposed upon me by Unraid. I have tried serveral things in OpenWRT (masquerading, forwarding...), all to no avail. I also tried on the CLI to add my VPN subnet to the iptables for DOCKER-USER, but this solved nothing: iptables -I DOCKER-USER -s 10.10.10.0/24 -j ACCEPT. Being new to Unraid, I was wondering if anyone could explain the mechanisms behind this blocking and where or how we are supposed to influence it? I find going in via iptables a little bit dark and cumbersome (coz I don't understand it haha). Wish ufw were available, or a GUI. Are we not supposed to manage networking at the firewall layer as Unraid users? That could be worrying, coz I like to think I have control ;). Thanks.

I filed a bug in the topic below. Not much reaction from the developers of UnRAID until now. I think your problem can be related to ours.

On 10/21/2025 at 2:14 PM, RJ59 said:

I filed a bug in the topic below. Not much reaction from the developers of UnRAID until now. I think your problem can be related to ours.

Today I fixed my problem. So, brief recap, in my case the problem existed on the far left on wireguard clients:

[wireguard on laptop] -> internet -> [wireguard on openwrt] -> lan -> [unraid host docker container:port]

Following some intense online searching and pondering... I wanted to add specifically the route for my wireguard vpn clients via the openwrt router on my unraid host:

ip route add 10.20.20.0/24 via 192.168.1.1

RTNETLINK answers: File exists

Unraid complains that that route already exists.

Looking at my docker networks I realized one was configured with the same subnet as my wireguard vpn clients on openwrt, presumably causing routing issues.

The solution: create a docker network using any other subnet than in my case wireguard 10.20.20.0/24, then connect containers to this network instead.

From this point, I was able to reach them from my vpn clients. 😮‍💨 One more thing learned.

For a network expert this probably 101, but I never knew docker networking has consequences beyond the host. One has to consider the entire network at home, not just on the host.