- What's New
- Product Bulletin
- Service Overview
- Billing
- Getting Started
-
User Guide
-
UCS Clusters
- Overview
- Huawei Cloud Clusters
-
On-Premises Clusters
- Overview
- Service Planning for On-Premises Cluster Installation
- Registering an On-Premises Cluster
- Installing an On-Premises Cluster
- Managing an On-Premises Cluster
- Attached Clusters
- Multi-Cloud Clusters
- Single-Cluster Management
- Fleets
-
Cluster Federation
- Overview
- Enabling Cluster Federation
- Using kubectl to Connect to a Federation
- Upgrading a Federation
-
Workloads
- Workload Creation
-
Container Settings
- Setting Basic Container Information
- Setting Container Specifications
- Setting Container Lifecycle Parameters
- Setting Health Check for a Container
- Setting Environment Variables
- Configuring a Workload Upgrade Policy
- Configuring a Scheduling Policy (Affinity/Anti-affinity)
- Configuring Scheduling and Differentiation
- Managing a Workload
- ConfigMaps and Secrets
- Services and Ingresses
- MCI
- MCS
- DNS Policies
- Storage
- Namespaces
- Multi-Cluster Workload Scaling
- Adding Labels and Taints to a Cluster
- RBAC Authorization for Cluster Federations
- Image Repositories
- Permissions
-
Policy Center
- Overview
- Basic Concepts
- Enabling Policy Center
- Creating and Managing Policy Instances
- Example: Using Policy Center for Kubernetes Resource Compliance Governance
-
Policy Definition Library
- Overview
- k8spspvolumetypes
- k8spspallowedusers
- k8spspselinuxv2
- k8spspseccomp
- k8spspreadonlyrootfilesystem
- k8spspprocmount
- k8spspprivilegedcontainer
- k8spsphostnetworkingports
- k8spsphostnamespace
- k8spsphostfilesystem
- k8spspfsgroup
- k8spspforbiddensysctls
- k8spspflexvolumes
- k8spspcapabilities
- k8spspapparmor
- k8spspallowprivilegeescalationcontainer
- k8srequiredprobes
- k8srequiredlabels
- k8srequiredannotations
- k8sreplicalimits
- noupdateserviceaccount
- k8simagedigests
- k8sexternalips
- k8sdisallowedtags
- k8sdisallowanonymous
- k8srequiredresources
- k8scontainerratios
- k8scontainerrequests
- k8scontainerlimits
- k8sblockwildcardingress
- k8sblocknodeport
- k8sblockloadbalancer
- k8sblockendpointeditdefaultrole
- k8spspautomountserviceaccounttokenpod
- k8sallowedrepos
- Configuration Management
- Traffic Distribution
- Observability
- Container Migration
- Pipeline
- Error Codes
-
UCS Clusters
- Best Practices
-
API Reference
- Before You Start
- Calling APIs
-
API
- UCS Cluster
-
Fleet
- Adding a Cluster to a Fleet
- Removing a Cluster from a Fleet
- Registering a Fleet
- Deleting a Fleet
- Querying a Fleet
- Adding Clusters to a Fleet
- Updating Fleet Description
- Updating Permission Policies Associated with a Fleet
- Updating the Zone Associated with the Federation of a Fleet
- Obtaining the Fleet List
- Enabling Fleet Federation
- Disabling Cluster Federation
- Querying Federation Enabling Progress
- Creating a Federation Connection and Downloading kubeconfig
- Creating a Federation Connection
- Downloading Federation kubeconfig
- Permissions Management
- Using the Karmada API
- Appendix
-
FAQs
- About UCS
-
Billing
- How Is UCS Billed?
- What Status of a Cluster Will Incur UCS Charges?
- Why Am I Still Being Billed After I Purchase a Resource Package?
- How Do I Change the Billing Mode of a Cluster from Pay-per-Use to Yearly/Monthly?
- What Types of Invoices Are There?
- Can I Unsubscribe from or Modify a Resource Package?
-
Permissions
- How Do I Configure Access Permissions for Each Function of the UCS Console?
- What Can I Do If an IAM User Cannot Obtain Cluster or Fleet Information After Logging In to UCS?
- How Do I Restore ucs_admin_trust I Deleted or Modified?
- What Can I Do If I Cannot Associate the Permission Policy with a Fleet or Cluster?
- How Do I Clear RBAC Resources After a Cluster Is Unregistered?
- Policy Center
-
Fleets
- What Can I Do If Cluster Federation Verification Fails to Be Enabled for a Fleet?
- What Can I Do If an Abnormal, Federated Cluster Fails to Be Removed from the Fleet?
- What Can I Do If an Nginx Ingress Is in the Unready State After Being Deployed?
- What Can I Do If "Error from server (Forbidden)" Is Displayed When I Run the kubectl Command?
- Huawei Cloud Clusters
- Attached Clusters
-
On-Premises Clusters
- What Can I Do If an On-Premises Cluster Fails to Be Connected?
- How Do I Manually Clear Nodes of an On-Premises Cluster?
- How Do I Downgrade a cgroup?
- What Can I Do If the VM SSH Connection Times Out?
- How Do I Expand the Disk Capacity of the CIA Add-on in an On-Premises Cluster?
- What Can I Do If the Cluster Console Is Unavailable After the Master Node Is Shut Down?
- What Can I Do If a Node Is Not Ready After Its Scale-Out?
- How Do I Update the CA/TLS Certificate of an On-Premises Cluster?
- What Can I Do If an On-Premises Cluster Fails to Be Installed?
- Multi-Cloud Clusters
-
Cluster Federation
- What Can I Do If the Pre-upgrade Check of the Cluster Federation Fails?
- What Can I Do If a Cluster Fails to Be Added to a Federation?
- What Can I Do If Status Verification Fails When Clusters Are Added to a Federation?
- What Can I Do If an HPA Created on the Cluster Federation Management Plane Fails to Be Distributed to Member Clusters?
- What Can I Do If an MCI Object Fails to Be Created?
- What Can I Do If I Fail to Access a Service Through MCI?
- What Can I Do If an MCS Object Fails to Be Created?
- What Can I Do If an MCS or MCI Instance Fails to Be Deleted?
- Traffic Distribution
- Container Intelligent Analysis
- General Reference
Copied.
gpu-device-plugin
Introduction
gpu-device-plugin is an add-on that supports GPUs in containers. If GPU nodes are used in the cluster, this add-on must be installed.
Constraints
- The driver to be downloaded must be a .run file.
- Only NVIDIA Tesla drivers are supported.
- When installing or reinstalling the add-on, ensure that the driver download address is correct and accessible. CCE does not verify the address validity.
- gpu-device-plugin enables you to download the driver and execute the installation script. The add-on status does not indicate whether the driver is installed successfully.
- Only T4 and V100 GPUs are supported.
- On-premises clusters support only Huawei Cloud EulerOS 2.0 (x86).
- Before adding GPU nodes to your cluster, ensure that libsecurec has been installed.
Prerequisites
- Add GPU nodes to your cluster. For details, see Adding Nodes to On-Premises Clusters.
- Label the nodes with accelerator: nvidia-{GPU model}. For details, see Adding Labels/Taints to Nodes.
Installing the Add-on
- Log in to the CCE console and click the cluster name to access the cluster console. In the navigation pane, choose Add-ons.
- Locate gpu-device-plugin in Add-ons Available and click Install.
- In the window that slides out from the right, configure the parameters as follows:
- Add-on Specifications: Select Default or Custom as required.
- Containers: This parameter can be configured only when Add-on Specifications is set to Custom.
- NVIDIA Driver: Use a driver address provided by CCE or enter the address of your custom NVIDIA driver. All GPU nodes in the cluster use the same driver.
GPU virtualization supports only GPU drivers 470.57.02, 510.47.03, and 535.54.03.
You are advised to use a driver address provided by CCE to match the driver version.NOTICE:
- If the download link is a public network address, for example, NVIDIA official website address (https://us.download.nvidia.com/tesla/470.103.01/NVIDIA-Linux-x86_64-470.103.01.run), associate EIPs with GPU nodes. For details about how to obtain the driver link, see Obtaining the Driver Link from Public Network.
- If the download link is an OBS URL, there is no need to bind an EIP to each GPU node. For details about how to obtain the driver link, see Obtaining the Driver Link from OBS.
- Ensure that the NVIDIA driver version matches the GPU node.
- If the driver version is changed, restart the node to apply the change.
CAUTION:
Before restarting a node, evict all pods on that node. Make sure to reserve GPU resources to avoid pod scheduling failures during node drainage. Insufficient resources can affect services.
- Use driver 470 or later for Huawei Cloud EulerOS 2.0 or Ubuntu 22.04 on which Linux Kernel 5.x is built.
Figure 1 Installing gpu-device-plugin
- Click Install.
Verifying the Add-on
After the add-on is installed, run the nvidia-smi command on the GPU node and the container that schedules GPU resources to verify the availability of the GPU and driver.
cd /usr/local/nvidia/bin &&./nvidia-smi
Container:
nvidia-smi
If GPU information is returned, the GPU is available and the add-on is successfully installed.
Obtaining the Driver Link from Public Network
- Log in to the CCE console.
- Click Create Node and select the GPU node to be created in the Specifications area. The GPU card model of the node is displayed in the lower part of the page.
- Log in to NVIDIA.
- Select the driver information on the NVIDIA Driver Downloads page, as shown in Figure 2. Operating System must be Linux 64-bit.
- After confirming the driver information, click SEARCH. A page is displayed, showing the driver information, as shown in Figure 3. Click DOWNLOAD.
- Obtain the driver link in either of the following ways:
- Method 1: As shown in Figure 4, find url=/tesla/470.103.01/NVIDIA-Linux-x86_64-470.103.01.run in the browser address box. Then, supplement it to obtain the driver link (https://us.download.nvidia.com/tesla/470.103.01/NVIDIA-Linux-x86_64-470.103.01.run). By using this method, you must associate EIPs with GPU nodes.
- Method 2: As shown in Figure 4, click Agree & Download to download the driver. Then, upload the driver to OBS and record the OBS URL. By using this method, you do not need to associate EIPs with GPU nodes.
Obtaining the Driver Link from OBS
- Upload the driver to OBS and set the driver file to public read. For details, see Uploading a File.
NOTE:
When the node is restarted, the driver will be downloaded and installed again. Ensure that the OBS bucket link of the driver is valid.
- Log in to the OBS console. In the navigation pane, select Object Storage.
- In the bucket list, click the bucket name you want. The Overview page of the bucket is displayed.
- In the navigation pane, choose Objects.
- Locate the target object and choose More > Copy Object URL to copy the driver link.
Figure 5 Obtaining the driver link
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot