Hi folks – Lately, my home assistant keeps losing network connection for no discernible reason. When it happens, I can’t access the web interface, and from the logs, it looks like HA can’t access any devices on the network either. From what I can tell, enp2s0 fails to renew its IP address.
I have a static IP address for HA in my wifi router, as well as put that same IP address in the web UI settings. All of this was working perfectly fine for a long time, so this issue is only recent (last few weeks).
The only error I see in the logs is for NetworkManager:
Mar 16 18:54:22 homeassistant NetworkManager[456]: <warn> [1742151262.4987] l3cfg[83255319a799acac,ifindex=2]: unable to configure IPv4 route: type unicast 10.0.0.0/24 dev 2 metric 100 mss 0 rt-src rt-kernel scope link pref-src 10.0.0.95
Mar 16 18:54:22 homeassistant NetworkManager[456]: <warn> [1742151262.4988] l3cfg[83255319a799acac,ifindex=2]: unable to configure IPv4 route: type unicast 0.0.0.0/0 via 10.0.0.1 dev 2 metric 20100 mss 0 rt-src dhcp pref-src 10.0.0.95
FWIW, systemctl restart NetworkManager immediately fixes the issue and gets me back on the network.
Any idea how can I go about isolating the problem?
My setup:
HAOS on VMWare Fusion, with bridged network that connects to my host (Mac OS) wifi (all of which worked perfectly fine up until a few weeks ago).
Noticed something strange - HA is bombarding DHCP on my wifi router multiple times a minute!
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:38:32
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:37:31
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:36:31
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:35:30
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:34:29
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:33:29
[DHCP IP: (10.0.0.95)] to MAC address 00:0C:29:61:E0:BE, Tuesday, Mar 18,2025 15:32:28
Looks like my static IP assignment in HA configuration hadn’t taken. I re-enabled it, and now I don’t see a DHCP request on my wifi router. But I am still surprised by the velocity of DHCP requests that were coming from HA!
Not a VM user but I would start there. HA won’t drop Ethernet connection unless it doesn’t exist. For this reason I would expect the VM has some issue. Again, guess.
I also forget if adapter will disappear if no tx/rx is available. Basically, if cable ethernet cable unplug is result the same. I forget. But this can be visually and software verified. Poor connection or other issues
As for disconnection, I tried disabling wifi on mac. It causes the IP route to disappear from HA. But when I re-enable wifi on mac, the route comes back in HA on its own, which means HAOS is robust to intermittent loss of network connectivity (if at all that is the problem) – which is great! I also tried disconnecting the network bridge on the VM, and then reconnecting it a few seconds later (to simulate random drops). Again, HA reconnects no problem.
Also, as I mentioned, systemctl restart NetworkManager immediately fixes the problem without me having to touch the VM settings at all. So I feel it has something to do with NetworkManager not doing its job properly, or something else inside HA that is causing NetworkManager to not work properly.
Yes. I believe problem is unrelated to HA or HAOS. It is likely VM/Host problem
HA is unlikely causing this. Look at host side problem, VM problem or bridge. Network manager untouched is stable so unless you made change to host it’s also not likely cause.