The iptablesHAOS dashboard isn't reachable on my local network, yet other applications function properly.
The iptablesHAOS dashboard isn't reachable on my local network, yet other applications function properly.
My configuration: Old machine using Mint + MATE as a server `.1.180`. On Wi-Fi it would be simpler. A VM is running HAOS (Home Assistant). I need to configure IPTables for network access. I have several IoT gadgets and another device with a Frigate server on the LAN. It worked before, but now it's not. What changed? Right now: - Phone connects to HAOS (:8123) which generates an IPv6 login error. IPv6 is turned off on HAOS. - Laptop connects to HAOS (:8123 with addons, hostname, both IPv4 and IPv6 enabled). - IoT devices communicate well with HAOS (MQTT, Frigate, RTSP, etc.). - Host connects to HAOS (:8123) and can access the UI directly from the VM's network at `.122.151`. - HAOS communicates with the web (downloads, etc.). - VM connects to devices (phone/laptop reachable via ping). - Host connects to devices (phone/laptop also reachable via ping). I tried routing all ports through the VM but it didn't help.
I suspect there might be an issue with DHCP or the VM's network connection. The problem seems to affect certain ports while leaving others unaffected.
The likely reason is that the port is active in the Wi-Fi IP range but not in the LAN range, indicating an old process is stuck or had an incomplete shutdown on the Wi-Fi side, or the firewall is blocking it. A reboot could fix the problem; otherwise, check for other causes. I’ve experienced similar issues before where a printer port was misused after an update, so I looked into the port and found the culprit. I disabled it since I no longer need it. Hope this helps.
You're concerned about the port issue, but you can check for activity with `sudo lsof -i :8123` on the host and VM. Regarding HAOS, it's using either IP:8123 or homeowner.local:8123.
I restarted the VM and the whole host. On the host machine I’m using `VM_IP:8123`. Other hosts trying `Host_IP:8123` or `host_hostname.local:8123` aren’t connecting. `Homeassistant.local:8123` fails because it’s inside the VM (except on the host, which works), so it uses the host’s name. Running `lsof` shows the connection is established from the host’s IP address.
Not very familiar with virtual machines, but I just set up Proxmox using an x99 salvage board and experimented with dual socket configurations. I currently have a single VM running there. This VM runs a Pi-hole service that receives its own IP from my router.
You're using QEMU/KVM on a minimal Linux system and adding HAOS as an extra layer.
I’m going to look up at that sail high above my head. What if I traced my route to my laptop instead of using ping? Honestly, I might feel pretty out of place right now. I’m only fairly good at virtual machines and networking—I pick up just enough to handle the immediate task, then lose most of it by tomorrow.
The VM lacks sufficient details, making tracert unavailable or irrelevant.
It seems you're adopting the most challenging method, which is acceptable. Consider switching your network from bridge to host or another configuration and observe the results. For a smoother experience, I’d suggest using Proxmox—built on Debian with a user-friendly interface. It offers better visibility into your networks and simplifies management. By the way, I’ve set up a virtual router inside a VM and connected it via a virtual LAN, which helps with organization. Good luck!