Note
Rebuild drivers with --dkms option if the Linux kernel is updated
[!NOTE]
Do not use Debian backports to install drivers. It does work, but the drivers are old, and it is significantly more work to remove the drivers.
https://docs.fileflows.com/guides/linux/proxmox-lxc-nvidia
echo -e "blacklist nouveau\noptions nouveau modeset=0" > /etc/modprobe.d/blacklist-nouveau.conf
update-initramfs -u
reboot
Install PVE Headers
apt install pve-headers-$(uname -r)
https://download.nvidia.com/XFree86/
wget -O NVIDIA-Linux-x86_64-525.60.13.run https://download.nvidia.com/XFree86/Linux-x86_64/525.60.13/NVIDIA-Linux-x86_64-525.60.13.run
chmod +x NVIDIA-Linux-x86_64-525.60.13.run
./NVIDIA-Linux-x86_64-525.60.13.run --check
./NVIDIA-Linux-x86_64-525.60.13.run
Say 'no' to 32-bit compatibility and to updating X-config
Add kernel module
echo -e '\n# load nvidia modules\nnvidia\nnvidia_uvm\nnvidia-drm\nnvidia-uvm' >> /etc/modules-load.d/modules.conf
Check cat /etc/modules-load.d/modules.conf. The following entries should be present:
GNU nano 5.4 /etc/modules-load.d/modules.conf * # /etc/modules: kernel modules to load at boot time.
#
# This file contains the names of kernel modules that should be loaded
# at boot time, one per line. Lines beginning with "#" are ignored.
# load nvidia modules
nvidia
nvidia_uvm
nvidia-drm
nvidia-uvm
Update Initramfs
update-initramfs -u -k all.
Note
This step may produce some warnings. They are unrelated to the driver install
Create rules to make relevant device at boot
nano /etc/udev/rules.d/70-nvidia.rules
KERNEL=="nvidia", RUN+="/bin/bash -c '/usr/bin/nvidia-smi -L && /bin/chmod 666 /dev/nvidia*'"
KERNEL=="nvidia_uvm", RUN+="/bin/bash -c '/usr/bin/nvidia-modprobe -c0 -u && /bin/chmod 0666 /dev/nvidia-uvm*'"
SUBSYSTEM=="module", ACTION=="add", DEVPATH=="/module/nvidia", RUN+="/usr/bin/nvidia-modprobe -m"
Install Nvidia Persistent Service to avoid drivers being unloaded when not in use
cp /usr/share/doc/NVIDIA_GLX-1.0/samples/nvidia-persistenced-init.tar.bz2 .
bunzip2 nvidia-persistenced-init.tar.bz2
tar -xf nvidia-persistenced-init.tar
Remove old service, if it exists
rm /etc/systemd/system/nvidia-persistenced.service
Install service
chmod +x nvidia-persistenced-init/install.sh
./nvidia-persistenced-init/install.sh
Check that it is OK
systemctl status nvidia-persistenced.service
rm -rf nvidia-persistenced-init*
Reboot
Check status
nvidia-smi
systemctl status nvidia-persistenced.service
ls -alh /dev/nvidia*
Shut down container. 195 and 508 come from ls -alh /dev/nvidia* and ls -alh /dev/dri*
nano /etc/pve/lxc/<vmid>.conf
lxc.cgroup2.devices.allow: c 195:* rwm
lxc.cgroup2.devices.allow: c 226:* rwm
lxc.cgroup2.devices.allow: c 237:* rwm
lxc.cgroup2.devices.allow: c 508:* rwm
lxc.mount.entry: /dev/nvidia0 dev/nvidia0 none bind,optional,create=file
lxc.mount.entry: /dev/nvidiactl dev/nvidiactl none bind,optional,create=file
lxc.mount.entry: /dev/nvidia-modeset dev/nvidia-modeset none bind,optional,create=file
lxc.mount.entry: /dev/nvidia-uvm dev/nvidia-uvm none bind,optional,create=file
lxc.mount.entry: /dev/nvidia-uvm-tools dev/nvidia-uvm-tools none bind,optional,create=file
Install Drivers on container
wget -O NVIDIA-Linux-x86_64-525.60.13.run https://download.nvidia.com/XFree86/Linux-x86_64/525.60.13/NVIDIA-Linux-x86_64-525.60.13.run
chmod +x NVIDIA-Linux-x86_64-525.60.13.run
./NVIDIA-Linux-x86_64-525.60.13.run --check
./NVIDIA-Linux-x86_64-525.60.13.run --no-kernel-module
Say no to 32-bit compatibility and X-conf
Reboot
root@UbuntuMEDIA:~# ls -alh /dev/nvidia*
crw-rw-rw- 1 root root 195, 254 Dec 17 21:09 /dev/nvidia-modeset
crw-rw-rw- 1 root root 508, 0 Dec 17 21:09 /dev/nvidia-uvm
crw-rw-rw- 1 root root 508, 1 Dec 17 21:09 /dev/nvidia-uvm-tools
crw-rw-rw- 1 root root 195, 0 Dec 17 21:09 /dev/nvidia0
crw-rw-rw- 1 root root 195, 255 Dec 17 21:09 /dev/nvidiactl
Install nvidia-container-tools
https://nvidia.github.io/libnvidia-container/ https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html
distribution=$(. /etc/os-release;echo $ID$VERSION_ID) \
&& curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
&& curl -s -L https://nvidia.github.io/libnvidia-container/experimental/$distribution/libnvidia-container.list | \
sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
Replace $distribution with ubuntu20.04 or most current distro
apt update && apt install -y nvidia-docker2
https://www.youtube.com/watch?v=Sypi9dfMLX0
In Portainer:
Container > Select Container > Duplicate/Edit > Runtime & Resources > Enable GPU
./NVIDIA-Linux-x86_64-525.60.13.run --uninstall
If installed with Debian Backports:
systemctl stop nvidia*
dpkg -l | grep -i nvidia
apt remove --purge '^nvidia-.*'
apt autoremove