tencent cloud

Cloud GPU Service

Release Notes and Announcements
Release Notes
Announcements
Product Introduction
Overview
Strengths
Scenarios
Notes
Instance Types
Computing Instance
Rendering Instance
Billing
Billing Overview
Renewal
Getting Started
User Guide
Logging In to Instances
Restarting Instances
Installing NVIDIA Driver
Uninstalling NVIDIA Driver
Upgrading NVIDIA Driver
Using GPU Monitoring and Alarm
Use Cases
Installing NVIDIA Container Toolkit on a Linux Cloud GPU Service
Using Windows Cloud GPU Service to build a Deep Learning Environment
Implementing Image Quality Enhancement with GN7vi Instances
Using Docker to Install TensorFlow and Set GPU/CPU Support
Using GPU Instance to Train ViT Model
Troubleshooting
GPU Instance Troubleshooting Guide
Troubleshooting Common Xid Errors
Collecting Log for GPU Instances
GPU Usage Shows 100%
VNC Login Failures
FAQs
Related Agreement
Special Terms for Committed Sales Model
Contact Us
문서Cloud GPU ServiceUse CasesInstalling NVIDIA Container Toolkit on a Linux Cloud GPU Service

Installing NVIDIA Container Toolkit on a Linux Cloud GPU Service

PDF
포커스 모드
폰트 크기
마지막 업데이트 시간: 2026-01-19 11:52:23

Scenarios

In Docker, containers cannot access the host GPU resources by default. To address this limitation, NVIDIA provides an official NVIDIA Docker container support solution that maps the host GPU runtime environment into containers. This solution has evolved through multiple stages—from the early nvidia-docker and nvidia-docker2 to today’s NVIDIA Container Toolkit.
NVIDIA Container Toolkit is a set of tools for container runtime environments that enables containers to use NVIDIA GPUs. This document describes how to install NVIDIA Container Toolkit on a Linux Cloud GPU Service using Tencent Cloud mirrors.

Procedure

Instance Environment Preparation

1. Check whether the GPU driver is installed.
Log in to the instance and run the following command to confirm whether the GPU driver is properly installed:
nvidia-smi
If GPU device information is returned normally, it indicates that the driver is installed; if the prompt nvidia-smi: command not found is returned, it indicates that the driver is not installed, and you need to refer to NVIDIA Driver Installation Guide to install the NVIDIA GPU driver.

2. Ensure Docker is installed and running.
Run the following command to verify Docker installation and that the Docker service is running:
docker ps
If the output is similar to the screenshot, Docker is installed and running. If not, refer to Building the Docker to install Docker.

3. Check whether NVIDIA Container Toolkit has not been installed.
Try to start the GPU container by running the following command:
docker run --gpus all [image name]
If you receive the error docker: Error response from daemon: could not select device driver "" with capabilities: [[gpu]], it indicates that NVIDIA Container Toolkit is not installed. The system lacks the required GPU runtime configuration and cannot recognize or allocate GPU resources to containers.

Install NVIDIA Container Toolkit

1. This section uses Tencent Cloud TencentOS Server, CentOS, RHEL, Rocky Linux, Ubuntu, and Debian GPU server environments as examples. For other operating systems, refer to Installing the NVIDIA Container Toolkit.
TencentOS Server/CentOS/RHEL/Rocky Linux
Ubuntu/Debian
 curl -s -L https://mirrors.tencentyun.com/libnvidia-container/stable/rpm/nvidia-container-toolkit.repo | sudo tee /etc/yum.repos.d/nvidia-container-toolkit.repo #Configure the nvidia-container-toolkit source
 sed -i 's/nvidia.github.io/mirrors.tencentyun.com/g' /etc/yum.repos.d/nvidia-container-toolkit.repo #Modify the nvidia-container-toolkit source to Tencent Cloud source
 sudo yum install -y nvidia-container-toolkit #Install nvidia-container-toolkit using yum on TencentOS 2.4/CentOS 7/RHEL 7
 sudo dnf install -y nvidia-container-toolkit #Install nvidia-container-toolkit using dnf on TencentOS 3.1/TencentOS 4/CentOS 8/RHEL 8/RHEL 9/Rocky Linux 8/Rocky Linux 9
 sudo rpm -qa | grep nvidia-container-toolkit #Check whether NVIDIA Container Toolkit is installed successfully. If related software information of nvidia-container-toolkit is displayed, it indicates successful installation.
curl -fsSL https://mirrors.tencentyun.com/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \\
  && curl -s -L https://mirrors.tencentyun.com/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \\
    sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \\
    sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list #Configure the nvidia-container-toolkit source
sed -i -e '/experimental/ s/^#//g' /etc/apt/sources.list.d/nvidia-container-toolkit.list #Enable the experimental repository
sed -i 's/nvidia.github.io/mirrors.tencentyun.com/g' /etc/apt/sources.list.d/nvidia-container-toolkit.list #Modify the nvidia-container-toolkit source to Tencent Cloud source
sudo apt-get update #Update the package list
sudo apt-get install -y nvidia-container-toolkit #Install the NVIDIA Container Toolkit package
sudo dpkg -l | grep nvidia-container-toolkit #Check whether NVIDIA Container Toolkit is installed successfully. If related software information of nvidia-container-toolkit is displayed, it indicates successful installation.
2. Run the following command to restart the docker service for NVIDIA Container Toolkit to take effect.
sudo systemctl restart docker
이전 주제: Using GPU Monitoring and Alarm다음 주제: Using Windows Cloud GPU Service to build a Deep Learning Environment

도움말 및 지원

문제 해결에 도움이 되었나요?

피드백