File tree Expand file tree Collapse file tree 1 file changed +8
-8
lines changed
Expand file tree Collapse file tree 1 file changed +8
-8
lines changed Original file line number Diff line number Diff line change @@ -120,15 +120,15 @@ kubectl apply -f volcano-development.yaml
120120- containerd:
121121
122122``` Bash
123- ctr -n=k8s.io i import gpu-device-plugin .tar
124- ctr -n=k8s.io i import cuda-client-update .tar
123+ ctr -n=k8s.io i import gpu_device_plugin .tar
124+ ctr -n=k8s.io i import cuda_client_update .tar
125125```
126126
127127- docker:
128128
129129``` Bash
130- docker load -i gpu-device-plugin .tar
131- docker load -i cuda-client-update .tar
130+ docker load -i gpu_device_plugin .tar
131+ docker load -i cuda_client_update .tar
132132```
133133
134134安装完成之后,执行如下命令将` client-update ` (cuda劫持)、` gpu-device-plugin ` (设备插件)安装部署yaml打包为helm chart包:
@@ -488,22 +488,22 @@ docker load -i openEuler-docker.x86_64.tar.xz
488488在` docker-build/client-update ` 目录下创建文件夹` mkdir GPU-client ` ,然后执行:
489489
490490``` Bash
491- docker build -t cuda-client-update :2.0 .
491+ docker build -t cuda_client_update :2.0 .
492492```
493493
494494在` docker-build/gpu-device-plugin ` 目录下执行:
495495
496496``` Bash
497- docker build -t gpu-device-plugin :2.0 .
497+ docker build -t gpu_device_plugin :2.0 .
498498```
499499
500500(上述代码的` . ` 不能忽略)
501501
502502至此,镜像制作完成,可以使用如下命令将镜像保存到本地:
503503
504504``` Bash
505- docker save -o gpu-device-plugin .tar gpu-device-plugin :2.0
506- docker save -o cuda-client-update .tar cuda-client-update :2.0
505+ docker save -o gpu_device_plugin .tar gpu_device_plugin :2.0
506+ docker save -o cuda_client_update .tar cuda_client_update :2.0
507507```
508508
509509volcano调度的打包流程与上述相仿,将编译产物复制到对应的` docker-build/ ` 所在的文件夹下:
You can’t perform that action at this time.
0 commit comments