ASUS Chromebox CN60 unreachable after upgrading to OS 16

[chrisgorgo]

Describe the issue you are experiencing

I cannot connect to my HA box after update (getting ERR_CONNECTION_RESET). Surprisingly some automations still work.

What operating system image do you use?

generic-x86-64 (Generic UEFI capable x86-64 systems)

What version of Home Assistant Operating System is installed?

16.0

Did the problem occur after upgrading the Operating System?

Yes

Hardware details

Chrome Box bare metal installation

Steps to reproduce the issue

1. Upgrade to 16
2. Try to access http://homeassistant.local:8123/

Anything in the Supervisor logs that might be useful for us?

Cannot access.

Anything in the Host logs that might be useful for us?

n/a

System information

No response

Additional information

No response

[chrisgorgo]

The box has wired connection and static IP.

[uranioh]

Same here.

[domectrl]

Seems like a lot of people with trouble when upgrading to 16. I have an issue as well, might be related. After the upgrade, my RPi5 does not use the dt_overlays that I had specified in /mnt/boot/config.txt. When I check the config with vi (using ssh to port 22222) the config is still there, but the kernel module that should be loaded (pwm_bcm2835) is no longer loaded with version 16. When I force loading with modprobe, the module is loaded, but still the devices are not created as in v. 15.2. I don't know, but this might be related to all issues people experience...

[vlad36N]

Same here. RPi4 has wired connection and static IP.

[Krojack]

I had to restore a backup of my VM.

Mine starts up and runs for about 15 minutes before the web interface just stops working. I check the console on the VM and it responds fine... my HA log is flooded with the following...

OSError: [Errno 24] No file descriptors available
2025-07-10 18:25:40.172 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None)
Traceback (most recent call last):
  File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection
    conn, addr = sock.accept()
                 ~~~~~~~~~~~^^
  File "/usr/local/lib/python3.13/socket.py", line 295, in accept
    fd, addr = self._accept()
               ~~~~~~~~~~~~^^

The log file grew to 1.3 gigs as well.

Some more entries from the log that I haven't seen before this. I've censored out my personal domain name in some spots.

2025-07-10 17:46:15.398 WARNING (SyncWorker_6) [urllib3.connectionpool] Connection pool is full, discarding connection: pve1.**********. Connection pool size: 10

2025-07-10 18:24:07.990 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None)
Traceback (most recent call last):
  File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection
  File "/usr/local/lib/python3.13/socket.py", line 295, in accept
OSError: [Errno 24] No file descriptors available
2025-07-10 18:24:07.991 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None)
Traceback (most recent call last):
  File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection
  File "/usr/local/lib/python3.13/socket.py", line 295, in accept
OSError: [Errno 24] No file descriptors available
2025-07-10 18:24:07.991 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None)
Traceback (most recent call last):
  File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection
  File "/usr/local/lib/python3.13/socket.py", line 295, in accept

[chrisgorgo]

I managed to fix this in my system. I logged in using remote access (via internet) and then enabled wifi interface in system settings.

[FreshYop]

Same error here

[mikeb9x]

A temporary working solultion: Set static IP of eth0 in console. Example:

ha network update eth0 --ipv4-address 192.168.1.5/24 --ipv4-gateway 192.168.1.1 --ipv4-method static --ipv4-nameserver 192.168.1.1

Then check if it's connected:

ha network info

[MichalKalita]

I can connect to HO from a notebook (I have a static address in DHCP settings in the router), but HO cannot connect to any device in the network. I need to set up a static address in network settings, and all integrations start working. After a restart, I must set up network settings again. Looks like the network address is not used after a restart. Other parameters like DNS and gateway look OK in form.

OS 16.0, device HO/Green

[FreshYop]

A temporary working solultion: Set static IP of eth0 in console. Example:

ha network update eth0 --ipv4-address 192.168.1.5/24 --ipv4-gateway 192.168.1.1 --ipv4-method static --ipv4-nameserver 192.168.1.1

Then check if it's connected:

ha network info

Even with that I still have lot of bugs 🤔

[mikeb9x]

A temporary working solultion: Set static IP of eth0 in console. Example:

ha network update eth0 --ipv4-address 192.168.1.5/24 --ipv4-gateway 192.168.1.1 --ipv4-method static --ipv4-nameserver 192.168.1.1

Then check if it's connected:

ha network info

Even with that I still have lot of bugs 🤔

Then you need to go back to HaOS 15.2 or wait for 16.1. I'm only providing a temporary solution for 'no connection' issue that worked for me and few other people.

[myhrmans]

Same issue here on a fresh install.

[FreshYop]

How can i go back to 15.2 when 16.0 is already install ?

[Krojack]

How can i go back to 15.2 when 16.0 is already install ?

You can try the following if you can get to the command prompt. Never done it myself as I just restore a backup of my entire virtual machine.

ha os update --version 15.2

I've seen talk about it here: https://www.reddit.com/r/homeassistant/comments/1lvwuox/failed_upgrade_os_160/

[FreshYop]

Hum it seems to be risky I'll wait and hope a new version fixing this bug

[BenBlv]

Same here with the same exact problem

[Azel4231]

Same here on Pi5 4GB.

• After upgrade, HAOS 16 starts fine. But after that it's not accissible over the network. (others have reported boot loops, but that's not my problem)
• Downgrading to 15.2 does not work without a functioning network connection.
• I have had a static IP before, but tried setting it again manually as described above. The problem is that the interface does not get assigned an IP. DNS and gateway are set, but not the interface IP itself.

Will try upgrading the eeprom next.

Edit:

• eeprom update did not fix the problem (I used these instructions https://www.raspberrypi.com/documentation/computers/raspberry-pi.html#bootloader_update_stable ). Still no v4 IP.
• but eth interface has v6 IPs, will try accessing them nex

[jagko]

My installation uses interface eth0 with reserved DHCP-Address. But in "ha network info" this interface was method "disabled". ha network update eth0 --ipv4-method auto --ipv6-method disabled solved this problem.

[Azel4231]

Great advice, thanks! Turns out I had another device that used the PI5 IP. It probably took over when the PI5 restarted after the update. Now I know why you should manage static IPs in one place (DHCP Server).

[drcjb]

Same problem for me. Raspberry Pi 5 8GB, 32GB ssd and 500GB NvME card. Did the upgrade last Thursday evening on dev server (bare bones HA but identical hw spec) and all went well, so I upgraded HAOS to v16.0 from v15.2 on Prod server. Everything seemed ok untill Friday morning when HA was unresponsive, I could see my dashboards but not the HA UI pages like Developer, Settings etc. Some timed automations still worked.

Rebooted and got nothing. Investigation showed the static IP address has dissapeared, so loaded that as above. Rebooted and HA started. I looked at the logs but nothing obvious (I'm no expert) other than in the Supervisor log a years worth of data was missing!

Restored to HAOS v15.2 and all well. Will wait for next HAOS upgrade but be much more causious about upgrading.

[FreshYop]

My installation uses interface eth0 with reserved DHCP-Address. But in "ha network info" this interface was method "disabled". ha network update eth0 --ipv4-method auto --ipv6-method disabled solved this problem.

I've also try this, but the HA instance seem's very unstable. Lot of 404, HA not accessible from the internet or from local network, then after few minutes, it become accessible... and then inaccessible. Very strange. I hope a fix quickly, if i can help with any log message or debug, you can contact me.

[myhrmans]

I've also try this, but the HA instance seem's very unstable. Lot of 404, HA not accessible from the internet or from local network, then after few minutes, it become accessible... and then inaccessible. Very strange.

Experiencing the exact same thing. Current version is super unstable and not usable :( Also able to debug what ever possible.

I've tried:

nmcli
set ipv4.addresses 192.168.1.130/24
set ipv4.gateway 192.168.1.1

Results in an empty address field.

network update enp0s18 --ipv4-address 192.168.1.130/24 --ipv4-method static Same result.

Setting it through the UI -> same result.

I also removed all config files in /mnt/overlay/etc/NetworkManager/system-connections and rebooted. Retried the same commands without success.

[ptrtech1]

Thousands of these error messages: 2025-07-13 17:31:11.258 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None) Traceback (most recent call last): File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection File "/usr/local/lib/python3.13/socket.py", line 295, in accept OSError: [Errno 24] No file descriptors available 2025-07-13 17:31:11.259 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None) Traceback (most recent call last): File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection File "/usr/local/lib/python3.13/socket.py", line 295, in accept OSError: [Errno 24] No file descriptors available 2025-07-13 17:31:11.259 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None) Traceback (most recent call last): File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection File "/usr/local/lib/python3.13/socket.py", line 295, in accept OSError: [Errno 24] No file descriptors available 2025-07-13 17:31:11.259 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None) Traceback (most recent call last): File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection File "/usr/local/lib/python3.13/socket.py", line 295, in accept OSError: [Errno 24] No file descriptors available 2025-07-13 17:31:11.260 ERROR (MainThread) [homeassistant] Error doing job: socket.accept() out of system resource (None) Traceback (most recent call last): File "/usr/local/lib/python3.13/asyncio/selector_events.py", line 178, in _accept_connection File "/usr/local/lib/python3.13/socket.py", line 295, in accept OSError: [Errno 24] No file descriptors available

[agners]

I managed to fix this in my system. I logged in using remote access (via internet) and then enabled wifi interface in system settings.

@chrisgorgo so you've enabled WiFi and that fixed the issue? Since you initially said that the box has wired connection, with that your system has two connections to the network no?

What type of hardware are you using exactly?

[ptrtech1]

I managed to fix this in my system. I logged in using remote access (via internet) and then enabled wifi interface in system settings.

@chrisgorgo so you've enabled WiFi and that fixed the issue? Since you initially said that the box has wired connection, with that your system has two connections to the network no?

What type of hardware are you using exactly?

FYI, as a point of reference, I'm dual homed (bridged) between two subnets with Ethernet on both (no WiFi at all).

[agners]

@Krojack @ptrtech1 the original post and most issues reported in here seem to be network related, where seemingly HAOS/NetworkManager lost connection information. Your issue look more like you are hitting the open file descriptor limits, this is tracked with #4164.

[myhrmans]

I managed to fix this in my system. I logged in using remote access (via internet) and then enabled wifi interface in system settings.

@chrisgorgo so you've enabled WiFi and that fixed the issue? Since you initially said that the box has wired connection, with that your system has two connections to the network no?

What type of hardware are you using exactly?

@agners Hit me up on discord if you wanna live debug, I'm in the home Assistant Discord.

@myhrmans is my tag

[chrisgorgo]

@chrisgorgo so you've enabled WiFi and that fixed the issue? Since you initially said that the box has wired connection, with that your system has two connections to the network no?

Correct: ethernet and wifi. Ethernet was working, and then I just neede to update wifi config (which I was able to do via remote connection).

What type of hardware are you using exactly?

ASUS Chromebox CN60

[shalak]

Same issue here. I have 2 NICs in my HassOS, "main" and "iot" - looks like the DHCP for "main" one did not assign the IP address (it did assign DNS and gateway though).

[FreshYop]

If anyone want to go back to previous version : ha os boot-slot other

I've try it and now it seem to be ok. You can test with : ha os info to see if you still have the 15.2 version.