NVIDIA Container Toolkit インストール2021/04/13 |
|
NVIDIA Container Toolkit をインストールすると、コンテナーから GPU を容易に利用可能となります。
|
|
| [1] | |
| [2] | |
| [3] | NVIDIA Container Toolkit をインストールします。 |
|
[root@dlp ~]# curl https://nvidia.github.io/nvidia-docker/centos8/nvidia-docker.repo > /etc/yum.repos.d/nvidia-docker.repo [root@dlp ~]# dnf -y install nvidia-container-toolkit |
| [4] | SELinux を有効にしている場合は、ポリシーの変更が必要です。 |
|
[root@dlp ~]#
vi nvidiasmi.te # 以下の内容で新規作成
module nvidiasmi 1.0;
require {
type container_runtime_tmpfs_t;
type container_t;
type xserver_misc_device_t;
class file { open read };
class chr_file { getattr ioctl open read write };
}
#============= container_t ==============
allow container_t container_runtime_tmpfs_t:file { open read };
allow container_t xserver_misc_device_t:chr_file { getattr ioctl open read write };
checkmodule -m -M -o nvidiasmi.mod nvidiasmi.te [root@dlp ~]# semodule_package --outfile nvidiasmi.pp --module nvidiasmi.mod [root@dlp ~]# semodule -i nvidiasmi.pp |
| [5] | コンテナーからの [nvidia-smi] コマンドの利用方法です。 |
|
# Pull 可能な Cuda イメージのタグ確認 [root@dlp ~]# curl -s https://registry.hub.docker.com/v1/repositories/nvidia/cuda/tags | sed "s/,/\n/g" | grep name "name": "10.0-base"} "name": "10.0-base-centos6"} "name": "10.0-base-centos7"} "name": "10.0-base-ubi7"} "name": "10.0-base-ubuntu14.04"} "name": "10.0-base-ubuntu16.04"} "name": "10.0-base-ubuntu18.04"} ..... ..... # Cuda 11.0 コンテナーイメージを Pull して [nvidia-smi] [root@dlp ~]# podman run -e NVIDIA_VISIBLE_DEVICES=all nvidia/cuda:11.0-base nvidia-smi Tue Apr 13 05:04:42 2021 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 460.32.03 Driver Version: 460.32.03 CUDA Version: 11.2 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 GeForce GTX 1070 Off | 00000000:05:00.0 Off | N/A | | 27% 35C P5 25W / 180W | 0MiB / 8119MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | No running processes found | +-----------------------------------------------------------------------------+ # 対話セッションから [nvidia-smi] [root@dlp ~]# podman run -it -e NVIDIA_VISIBLE_DEVICES=all nvidia/cuda:11.0-base bash root@39eec91a01d8:/# nvidia-smi
Tue Apr 13 05:05:16 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.32.03 Driver Version: 460.32.03 CUDA Version: 11.2 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 GeForce GTX 1070 Off | 00000000:05:00.0 Off | N/A |
| 27% 35C P5 18W / 180W | 0MiB / 8119MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
root@39eec91a01d8:/# exit
[root@dlp ~]# podman images REPOSITORY TAG IMAGE ID CREATED SIZE docker.io/nvidia/cuda 11.0-base 2ec708416bb8 7 months ago 125 MB |
| Sponsored Link |