/var/log is getting full (currently 100 % used) - when running vm's

[Baitham]

This happened to me yesterday when I launched a windows VM that I have been using for about 3 months now. I am running 2 windows vm's both seemed to fill up the log and shut down. 

After searching online and doing some trouble shooting-with no luck- I decided to delete all vm's,vdisk, libvirt.image and starting over. 

The problem did not happen when setting up my first windows 10 vm which I only use though vnc.

The second windows vm with gpu passthrough is where I get the problem so I'm guessing it has to do with that. So I tried to reinstall NVidia driver plugin updated and made sure it sees my GPU 1070ti. 

After a reboot all returns to normal. the issue only comes back when that VM is turned on. 

Im running the newest unraid os build everything was running fine after update.

 

 

I'm attaching the diagnostics file here. 

 

tower-diagnostics-20210324-1943.zip

Edited March 25, 2021 by Baitham

[John_M]

I suspect this is your problem:

 

Mar 24 19:32:27 Tower kernel: vfio-pci 0000:01:00.0: BAR 3: can't reserve [mem 0xa0000000-0xa1ffffff 64bit pref]
### [PREVIOUS LINE REPEATED 1115022 TIMES] ###

 

That's a lot of error messages, no wonder the log space is full!

 

You have this:

 

01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP104 [GeForce GTX 1070 Ti] [10de:1b82] (rev a1)
	Subsystem: Gigabyte Technology Co., Ltd GP104 [1458:3794]
	Kernel driver in use: nvidia
	Kernel modules: nvidia_drm, nvidia

 

You say you're trying to pass through the GPU to a VM. You can't do that if it's already bound to a driver in Unraid. You need to remove the Nvidia driver.