My server has been running fine for over a year now, but recently ive been having daily crashes and I cannot figure out the reason…
The system will become unresponsive, and if i check the console window I see that.
I have tried:
same result. Has anyone seen this before?
What OS are you running virtualbox on ?
Win10 pro on i7 6800k, 32gb
remove hyper-v and prevent it from running on the host.
cmd as admin then:
bcdedit /set hypervisorlaunchtype off
DISM /Online /Disable-Feature:Microsoft-Hyper-V
then restart it.
Will try this now
Any updates to this? You need to tell us if it worked or not for your case.
It has not crashed this morning, but I need to give it more time to see for sure.
Well, no crash again so far. So I think disabling Hyper V may have solved the issue.
Figured as much since if you are not using hyper-v and its enabled it conflicts with other virtual appliance software.
I am having the same issue even if i disabled the Hyper V. Here are the messages i see:
I have read in another topic that this is caused by cores time sync. They suggest to run HA only with one core for Virtual Box. However, even in the documentation is stated to give it at least 2 cores. I tried reducing my initial 4 cores to 2 but it didn’t fix the problem.
Anyone else?
This is my current configuration in the VM settings for my instance running off my Win11 Pro installed NUC5i7RYH:
Thanks. Here is mine:
From what i can see, you are using Nested Paging, PAE/NX, and KVM paravirtualization while mine has only the latest.
While searcing on it on internet, i found the following thread:
RCU preempt cpu Stalls , able to solve it by disabling GPU - NXP Community the OP mentioned that he removed any HDMI devices and disabled GPU.
With that in mind, i am using VNC on the machine that Virtual Box is installed and as you may know, VNC requires a connected active monitor in order to connect remotely. For that reason, i have one of these HDMI Dummy plugs in order to be able to connect:
Maybe this is what is causing the issue. So a few things to try:
Thanks for your help
I use ultravnc to connect to the nuc from the desktop when I need to remote into locally and anydesk when I am not home for both the nuc and main pc.
The nuc is plugged into a monitor which is vesa mounted to my rack atm as I also run synology surveliance center on it for my camera feeds and have hass agent installed to control the spotify app volume automatically when I use it for white noise to get sleep.
The nuc is inside an akasa passive cooled case so the only noise I really get from the rack is the platter drives in the nas when they get accessed.
I used one of those headless mode plugs when I had my spare pi3 setup at my parents place to remote into via anydesk (got slow so its in my box of gear again after getting remote home assistant instance setup to try it as a bluetooth proxy for the main instance).
(I may get carried away with my setup explanations but if it helps overall then I am doing my job helping give ideas heh).
also to allow it to auto start in headless mode I just add this shortcut to taskchedular to run at bootup of the nuc even with the user account not being signed in:
https://jez.me/article/how-start-virtualbox-vm-headless-windows-10
The skyconnect is set in the usb devices filter with the port 001.
To start automatically the VM you can also go to VirtualBox, right click on your machine and select “Create a shortcut on Desktop”. Then you can place this shortcut on the startup folder (shell:startup). No need to create an additional task for this.
I have changed my settings to be similar to yours and i will give it some time to see if it’s gonna throw the messages! Will report back!
It’s what I have done, I just added the taskschedular as a failsafe to make sure it boots up using the shortcut without login to the host so when host gets turned off or restarted it is back online with it reliably.
Speaking off I also have a dedicated button on my streamdeck xl to manually restart the HA instance when I make config changes or updates request it.
Let us know how stable your instance is when using similar when ready to check in.
So, i have an update.
I tried to match the configuration shared above by TH3xR34P3R and also disconnect the dummy HDMI adapter i have with no positive results.
The only things that seems to mitigate the issue is when i reduce the cores to 1. From performance perspective i didn’t see a big difference so for now i will keep it with 1 core, but this is something that needs to be fixed.
I’ve made a little change with the setup as compared to the previous post:
I increased the ram allocation to 4096 from 2048, removed everything but the hard disk for the boot order and renamed the vdi image to just haos whilst changing it to be detected as an SSD instead of a HDD in the controller settings when mounting it again just to clean it up and since its running off a SSD in the nuc.
Otherwise everything else is the same as before. Figured I would post this version just to have it shown and available for others.
For reference I am using this build of virtual box at this time:
This solved my problem!
