add GPU docs #706

wbc6080 · 2025-06-23T06:52:42Z

Please check if the PR fulfills these requirements

The commit message follows our guidelines
Tests for the changes have been added (for bug fixes / features)
Docs have been added / updated (for bug fixes / features)

Which issue(s) this PR fixes:

Fixes #

What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)

docs update

What is the new behavior (if this is a feature change)?

Added documentation for managing edge GPU nodes and introduce how to use GPU resources in edge applications.

kubeedge-bot · 2025-06-23T06:52:47Z

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
To complete the pull request process, please assign fisherxu after the PR has been reviewed.
You can assign the PR to them by writing /assign @fisherxu in a comment when ready.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

fujitatomoya

overall lgtm, this would be useful.

fujitatomoya · 2025-06-23T21:55:17Z

docs/advanced/gpu.md

@@ -0,0 +1,296 @@
+---
+title: Edge Pods use GPU


looks like this section dedicates for the Nvidia GPU, how to bind that to the application container. in that case, it probably would be better to clearly add it to the section title?

Suggested change

title: Edge Pods use GPU

title: Edge Pods use Nvidia GPU

fujitatomoya · 2025-06-23T22:03:24Z

docs/advanced/gpu.md

+   root@edgenode:~/release-v1.16.0-rc.1-experimental/packages/ubuntu18.04/amd64# ls
+   libnvidia-container1_1.16.0~rc.1-1_amd64.deb      libnvidia-container-tools_1.16.0~rc.1-1_amd64.deb      nvidia-container-toolkit-operator-extensions_1.16.0~rc.1-1_amd64.deb
+   libnvidia-container1-dbg_1.16.0~rc.1-1_amd64.deb  nvidia-container-toolkit_1.16.0~rc.1-1_amd64.deb
+   libnvidia-container-dev_1.16.0~rc.1-1_amd64.deb   nvidia-container-toolkit-base_1.16.0~rc.1-1_amd64.deb


indent is off?

Suggested change

root@edgenode:~/release-v1.16.0-rc.1-experimental/packages/ubuntu18.04/amd64# ls

libnvidia-container1_1.16.0~rc.1-1_amd64.deb libnvidia-container-tools_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit-operator-extensions_1.16.0~rc.1-1_amd64.deb

libnvidia-container1-dbg_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit_1.16.0~rc.1-1_amd64.deb

libnvidia-container-dev_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit-base_1.16.0~rc.1-1_amd64.deb

root@edgenode:~/release-v1.16.0-rc.1-experimental/packages/ubuntu18.04/amd64# ls

libnvidia-container1_1.16.0~rc.1-1_amd64.deb libnvidia-container-tools_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit-operator-extensions_1.16.0~rc.1-1_amd64.deb

libnvidia-container1-dbg_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit_1.16.0~rc.1-1_amd64.deb

libnvidia-container-dev_1.16.0~rc.1-1_amd64.deb nvidia-container-toolkit-base_1.16.0~rc.1-1_amd64.deb

fujitatomoya · 2025-06-23T22:03:33Z

docs/advanced/gpu.md

+Execute the following command in this directory to complete the installation:
+
+```shell
+   sudo apt install ./*


Suggested change

sudo apt install ./*

sudo apt install ./*

Signed-off-by: wbc6080 <[email protected]> Co-authored-by: ming.tang <[email protected]>

kevin-wangzefeng · 2025-07-15T02:47:00Z

/gemini review

gemini-code-assist

Code Review

This pull request introduces documentation for configuring and using GPUs in KubeEdge. The documentation covers setting up the GPU environment, managing GPU nodes, and testing GPU allocation. The documentation is provided in both English and Chinese. The review focuses on improving clarity and flow in both languages.

gemini-code-assist · 2025-07-15T02:48:54Z

docs/advanced/gpu.md

+For specific installation guides, please refer to [Container Runtime](https://kubeedge.io/docs/setup/prerequisites/runtime)
+
+:::tip
+Since KubeEdge v1.14, support for Dockershim has been removed, and use Docker runtime to manage edge containers is no longer supported. If you still need to use Docker, you need to install [cri-dockerd](https://kubeedge.io/docs/setup/prerequisites/runtime#docker-engine) after installing Docker.


Consider rephrasing to improve clarity: "Since KubeEdge v1.14, direct Docker runtime support for managing edge containers has been removed. If you still need to use Docker, install cri-dockerd after installing Docker."

Suggested change

Since KubeEdge v1.14, support for Dockershim has been removed, and use Docker runtime to manage edge containers is no longer supported. If you still need to use Docker, you need to install [cri-dockerd](https://kubeedge.io/docs/setup/prerequisites/runtime#docker-engine) after installing Docker.

Since KubeEdge v1.14, direct Docker runtime support for managing edge containers has been removed. If you still need to use Docker, install cri-dockerd after installing Docker.

gemini-code-assist · 2025-07-15T02:48:54Z

docs/advanced/gpu.md

+keadm join --cgroupdriver=systemd \
+	--cloudcore-ipport="THE-EXPOSED-IP":10000 \
+	--kubeedge-version=v1.17.0 \
+	--token="YOUR TOKEN"
+	--remote-runtime-endpoint=unix:///var/run/cri-dockerd.sock


The command examples for Docker and Containerd could benefit from a brief explanation of the parameters used, such as --cgroupdriver, --cloudcore-ipport, and --remote-runtime-endpoint. This would help users understand the purpose of each parameter.

gemini-code-assist · 2025-07-15T02:48:54Z

docs/advanced/gpu.md

+
+```shell
+# After deployment, check whether it is successfully deployed on the edge node
+[root@master-01 ~]# kubectl get daemonsets.apps -n kube-system|grep nvidia


Consider adding a link to the KubeEdge documentation that explains how to deploy applications on edge nodes. This would provide users with a more complete understanding of the process.

gemini-code-assist · 2025-07-15T02:48:54Z

i18n/zh/docusaurus-plugin-content-docs/current/advanced/gpu.md

+- 如果边缘节点能够直接访问外部网络，可以直接按照[官方文档](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html)进行安装
+- 如果边缘节点无法直接访问外部网络，则需要在网络可以联通的机器上下载[官方离线安装包](https://github.com/NVIDIA/nvidia-container-toolkit/releases)，将安装包传入边缘节点完成解压。
+  解压后目录中应该出现如下的文件：


Consider rephrasing for better flow: "如果边缘节点无法直接访问外部网络，则需要在可以联通网络的主机上下载官方离线安装包，然后将安装包传入边缘节点并解压。"

Suggested change

- 如果边缘节点能够直接访问外部网络，可以直接按照[官方文档](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html)进行安装

- 如果边缘节点无法直接访问外部网络，则需要在网络可以联通的机器上下载[官方离线安装包](https://github.com/NVIDIA/nvidia-container-toolkit/releases)，将安装包传入边缘节点完成解压。

解压后目录中应该出现如下的文件：

如果边缘节点无法直接访问外部网络，则需要在可以联通网络的主机上下载[官方离线安装包](https://github.com/NVIDIA/nvidia-container-toolkit/releases)，然后将安装包传入边缘节点并解压。

gemini-code-assist · 2025-07-15T02:48:54Z

i18n/zh/docusaurus-plugin-content-docs/current/advanced/gpu.md

+
+这里以 Docker 和 Containerd 容器运行时作为示例：
+
+```shell


Consider rephrasing for better clarity: "这里以 Docker 和 Containerd 容器运行时为例："

Suggested change

```shell

这里以 Docker 和 Containerd 容器运行时为例：

gemini-code-assist · 2025-07-15T02:48:54Z

i18n/zh/docusaurus-plugin-content-docs/current/advanced/gpu.md

+检查 k8s-device-plugin 是否成功部署：
+
+```shell
+# After deployment, check whether it is successfully deployed on the edge node


Consider rephrasing for better readability: "检查 k8s-device-plugin 是否成功部署："

Suggested change

# After deployment, check whether it is successfully deployed on the edge node

检查 k8s-device-plugin 是否成功部署：

kubeedge-bot requested review from edisonxiang and kevin-wangzefeng June 23, 2025 06:52

kubeedge-bot added the size/XL Denotes a PR that changes 500-999 lines, ignoring generated files. label Jun 23, 2025

fujitatomoya reviewed Jun 23, 2025

View reviewed changes

add GPU docs

6b8fd13

Signed-off-by: wbc6080 <[email protected]> Co-authored-by: ming.tang <[email protected]>

wbc6080 force-pushed the gpu-docs branch from a06c31f to 6b8fd13 Compare June 24, 2025 02:58

wbc6080 closed this Jun 26, 2025

wbc6080 reopened this Jun 26, 2025

gemini-code-assist bot reviewed Jul 15, 2025

View reviewed changes

	Since KubeEdge v1.14, support for Dockershim has been removed, and use Docker runtime to manage edge containers is no longer supported. If you still need to use Docker, you need to install [cri-dockerd](https://kubeedge.io/docs/setup/prerequisites/runtime#docker-engine) after installing Docker.
	Since KubeEdge v1.14, direct Docker runtime support for managing edge containers has been removed. If you still need to use Docker, install cri-dockerd after installing Docker.


		这里以 Docker 和 Containerd 容器运行时作为示例：

		```shell

	```shell
	这里以 Docker 和 Containerd 容器运行时为例：

	# After deployment, check whether it is successfully deployed on the edge node
	检查 k8s-device-plugin 是否成功部署：

add GPU docs #706

Are you sure you want to change the base?

add GPU docs #706

Conversation

wbc6080 commented Jun 23, 2025

Uh oh!

kubeedge-bot commented Jun 23, 2025

Uh oh!

fujitatomoya left a comment

Choose a reason for hiding this comment

Uh oh!

fujitatomoya Jun 23, 2025

Choose a reason for hiding this comment

Uh oh!

fujitatomoya Jun 23, 2025

Choose a reason for hiding this comment

Uh oh!

fujitatomoya Jun 23, 2025

Choose a reason for hiding this comment

Uh oh!

wbc6080 Jun 24, 2025

Choose a reason for hiding this comment

Uh oh!

kevin-wangzefeng commented Jul 15, 2025

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants