HI All
I have been having an issue with my Home assistant setup. At random times during the day or at night ,the logs will show it is losing access to my wireless Lights and switches. When I check inside HA, I can see that they usually reconnect (Yeelights seem to take 30 seconds, TP-LINK lights and switches a usually about 6 seconds)
Before that a little background.
I was previously running my Home assistant on a Raspberry Pi 3. I decided to migrate to a NUC which is running proxmox.
The NUC is about 5 years old, I got it second hand off ebay.
Once proxmox was set up, I ran the Whiskerz007 script to install the virtual machine. Then I loaded a snapshot, restored, and set the IP address as the same as my old Pi. This was successful. However I did start to notice random disconnects from wireless devices pretty much straight away…
So as part of my troubleshooting I have done the following
At first I thought it was my pi-hole which runs on an old original Pi B which was causing issues cause it wasn’t able to handle it, So I moved all the clients off it to use the build in ADGuard. However this didn’t appear to do anything.
Installed a new VM with the Whiskerz007 script, restored snapshot, to see if it is my config. Both machine get disconnections at the same time
Installed a new VM with the Whiskerz007 script, and just set up a basic config connecting to my lights and switches. Also have disconnections at the same time
Installed a Linux Machine and installed supervisor on it, Set it up with a basic config to connect to the lights/switches. Still disconnecting
Change the network cable on my NUC. Still disconnecting
Move the Network cable to another port on my router . Still disconnecting
I am starting to think it is my NUC which is having the issue, however I cannot prove it.
One other issue related to this is my one tasmota light which connects to HA Mosquitto. Whenever this issue with disconnection happens. THe light turns off…I am not sure if there is a command which I can run on the tasmota console to have it stay on when it reconnects.
My next plan is to fire up an old desktop which I have, put proxmox onto it and see if it occurs with it…
So does anyone have any tricks or advise they could provide…
I have about 30 devices on my wireless. I am running a Netgear R7000. My first thought was that it was my network. However I have not increased the number of devices since I moved from the Pi to the NUC. Whilst I would get an occasional disconnect when I was using the Pi, it seems to be occurring more often since I am using the NUC
Well over the weekend, I spent my lockdown trying to troubleshoot this issue more.
First I changed my 2.4GHz channel from 13 to 6. Still experienced random dropouts
I turned off my port forwarding which allowed me to get access to my home network from external. The reason for this is I saw a lot of logs entries in my router which was been forwarded, to be Home Assistant, so I thought that it was been overloaded by external access, however that also was unsuccessful
Finally one which I thought was a bit of a stab in the dark, I removed the HDD which was sitting on the router, which was used as a NAS. This also did not work
So my next test was to fire up my old desktop, Install Proxmox, Run the Whiskerz007 script and then set up a basic config. This seems to show that some devices were staying connected on the desktop, but losing it on the NUC. So I did a full snapshot restore, and so far, it seems that the old desktop is working, whislt the NUC is still getting disconnects.
Below is the 2 sets of logs
NUC
Old desktop
So before I resell my NUC, I decided to get a usb3 to ethernet and see if it is the NIC
Only other thing I can’t check is the combee II, since it is still in the NUC, but I might try that on the desktop just to be sure. However I would be surprised if the combee II would be causing interference, since it was fine with my pi3
I had interference issues with the NUC and had to put the ConBee II on an USB extension cable, because the ConBee II stick can have issues due to the USB 3 port (Pi 3 has only USB 2). Intel published an article here, where you can read more about this.
HI Everyone
If anyone is following this topic or find it in the future, I have made some progress
I noticed on my proxmox server there was some error which occured at the same time my home assistant lost access to some of the devices
The error was e1000e eno1: Detected Hardware Unit Hang:
I did some googleing and it showed a topic on the proxmox forum here
Essentially the fix is to run this command on the proxmox console ethtook -K <interface> tso off gso off
So overnight, there was no issues with disconnecting…
So I was happy in the morning, However it didn’t help that a lot of my wireless were not connecting to the network in the morning. I concluded it was the router, as I couldn’t ping the device, so I rebooted my router, and they appeared back…
Will monitor it for a bit, but this is hopefully the issue.