Issues & Solve

Last Update: 21/11/20

CUDA out of memory

• Kill dead processes;

# replace '*' with 0, 1, ..., find running $PID
lsof /dev/nvidia*
kill -9 $PID

• reload Nvidia driver;

# Reload without rebooting
# close display management
systemctl isolate multi-user.target

# disable nvidia driver（if returns relying on xxx，rmmod them as well）
rmmod nvidia

# load nvidia driver
modprobe nvidia

# restart display management
systemctl start graphical.target

Or

sudo apt install nvidia-modprobe

• Reboot.

---

Docker Run without Root

sudo chmod a+rw /var/run/docker.sock

---

Docker Nvidia Runtime

Install Nvidia Container Runtime

curl -s -L https://nvidia.github.io/nvidia-container-runtime/gpgkey | \
  sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-container-runtime/$distribution/nvidia-container-runtime.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-runtime.list
sudo apt-get update
sudo apt-get install nvidia-container-runtime

Modify Config

sudo vim /etc/docker/daemon.json

# /etc/docker/daemon.json
{
    "runtimes": {
        "nvidia": {
            "path": "nvidia-container-runtime",
            "runtimeArgs": []
        }
    },
    "default-runtime": "nvidia"
}

Reload docker

sudo systemctl daemon-reload
sudo systemctl restart docker

---

Mount WebDav as 777

sudo apt install davfs2
sudo mount -t davfs http://your-url-of-webdav/ /your/mount/point/path -o uid=$YOURUID,gid=$YOURGID,dir_mode=0777,file_mode=0777

---

Mount Disk as 777

# find target disk, like /dev/sda or /dev/sda1
sudo fdisk -l

# mount once. file permission is (777-umask)
sudo mount /dev/your-disk /your/mount/point/path -o umask=000

# mount as startup
# find $UID, like 3e395c6c-8e04-4bd6-93cd-1152fa797262
sudo blkid /dev/your-disk
sudo vim /etc/fstab

# /etc/fstab
# add line
UID=$UID /your/mount/point/path ext4 auto,exec,user,rw,umask=000 0 0