Kylin Linux Advanced Server V10 (Tercel) aarch64 安装nvidia-container-runtime docker nvidia-smi

Installer nvidia-container-runtime

yumdownloader télécharge le package rpm correspondant à centos8 AArch6 et installe :

[root@ceph3 libnvidia-container-rpms]# yum install ./*aarch64.rpm
Kylin Linux Advanced Server 10 - Os                                                                         133 MB/s | 3.8 MB     00:00    
Last metadata expiration check: 0:00:02 ago on 2023 8-р сар 08, Мя 10:29:25.
Dependencies resolved.
============================================================================================================================================
 Package                                         Architecture              Version                     Repository                      Size
============================================================================================================================================
Installing:
 libnvidia-container1                            aarch64                   1.13.5-1                    @commandline                   930 k
 libnvidia-container-tools                       aarch64                   1.13.5-1                    @commandline                    55 k
 nvidia-container-toolkit                        aarch64                   1.13.5-1                    @commandline                   797 k
 nvidia-container-toolkit-base                   aarch64                   1.13.5-1                    @commandline                   2.7 M

Transaction Summary
============================================================================================================================================
Install  4 Packages

Total size: 4.5 M
Installed size: 15 M
Is this ok [y/N]: y
Downloading Packages:
Running transaction check
Transaction check succeeded.
Running transaction test
Transaction test succeeded.
Running transaction
  Preparing        :                                                                                                                    1/1 
  Installing       : nvidia-container-toolkit-base-1.13.5-1.aarch64                                                                     1/4 
  Installing       : libnvidia-container1-1.13.5-1.aarch64                                                                              2/4 
  Running scriptlet: libnvidia-container1-1.13.5-1.aarch64                                                                              2/4 
  Installing       : libnvidia-container-tools-1.13.5-1.aarch64                                                                         3/4 
  Installing       : nvidia-container-toolkit-1.13.5-1.aarch64                                                                          4/4 
  Running scriptlet: nvidia-container-toolkit-1.13.5-1.aarch64                                                                          4/4 
  Verifying        : libnvidia-container1-1.13.5-1.aarch64                                                                              1/4 
  Verifying        : libnvidia-container-tools-1.13.5-1.aarch64                                                                         2/4 
  Verifying        : nvidia-container-toolkit-1.13.5-1.aarch64                                                                          3/4 
  Verifying        : nvidia-container-toolkit-base-1.13.5-1.aarch64                                                                     4/4 

Installed:
  libnvidia-container1-1.13.5-1.aarch64            libnvidia-container-tools-1.13.5-1.aarch64   nvidia-container-toolkit-1.13.5-1.aarch64  
  nvidia-container-toolkit-base-1.13.5-1.aarch64  

Complete!

construire la configuration

[root@ceph3 libnvidia-container-rpms]# nvidia-ctk runtime configure --runtime=docker
INFO[0000] Loading docker config from /etc/docker/daemon.json 
INFO[0000] Config file does not exist, creating new one 
INFO[0000] Wrote updated config to /etc/docker/daemon.json 
INFO[0000] It is recommended that the docker daemon be restarted. 

afficher le contenu

[root@ceph3 libnvidia-container-rpms]# cat /etc/docker/daemon.json 
{
    "runtimes": {
        "nvidia": {
            "args": [],
            "path": "nvidia-container-runtime"
        }
    }
}

redémarrer le menu fixe

systemctl restart docker

tester nvidia-smi

 En conséquence, la carte graphique Tesla T4 a été reconnue.

Les références:

Guide d'installation — documentation container-toolkit 1.13.5