检测到您已登录华为云国际站账号，为了您更好的体验，建议您访问国际站服务网站 https://www.huaweicloud.com/intl/zh-cn
不再显示此消息
检测到您已登录华为云国际站账号，为了您更好的体验，建议您访问国际站服务网站 https://www.huaweicloud.com/intl/zh-cn
不再显示此消息
the configuration of CCE AI Suite (NVIDIA GPU) in a cluster has been intrusively modified.
Failed to delete the server group of master Major Check whether the master node (ECS) is deleted from the cluster. Failed to delete the virtual IP for the master Major Check whether the virtual IP address is deleted from the cluster.
Make sure that the selected CIDR block includes either the container subnet (for Cloud Native Network 2.0) or the node subnet (for tunnel network and VPC network). Figure 5 Adding an SNAT rule Check whether the pod can access the Internet.
Configuring Intra-VPC Access This section describes how to access an intranet from a container (outside the cluster in a VPC), including intra-VPC access and cross-VPC access.
is used to transmit VXLAN packets in container networking (involved when the container tunnel network model is used) Dynamic port (30000-32767) TCP Listening port of kube-proxy for layer-4 load balancing.
Figure 2 Kubernetes cluster architecture Master node A master node is the machine where the control plane components run, including the API server, scheduler, controller manager, and etcd. API server: a transit station for components to communicate with each other.
Default: false Console/API Enabling audit log collection for resource query (get) and update (update) operations in the kube-system namespace kube-system-audit By default, audit logs do not record resource query (get) and update (update) operations in the kube-system namespace.
" } } } Example Responses Status code: 200 This indicates that the public API server address of the cluster is successfully bound.
Check whether the cluster network kernel component is abnormal (container tunnel network model: openswitch kernel component; VPC network model: ipvlan kernel component).
API & kubectl Why Is "Error from server (Forbidden)" Displayed When I Use kubectl? Image Repository How Do I Create a Docker Image and Solve the Problem of Slow Image Pull?
and a maximum of 50 worker nodes) - cce.dec.s1.medium: a medium-scale DeC CCE cluster with one control plane node and a maximum of 200 worker nodes) - cce.dec.s1.large: a large-scale DeC CCE cluster with one control plane node and a maximum of 1,000 worker nodes) - cce.dec.s1.xlarge
The remaining period is calculated as follows: Remaining period = 12 (Remaining days in April)/30 (Maximum number of days in April) + 8 (Remaining days in May)/31 (Maximum number of days in May) = 0.6581 Specifications modification price = (551.46 - 285.66) x 0.6581 = $174.92 USD
CBR cloud server backup Master node disks, including component images, configurations, logs, and etcd data One-click backup on a web page (manually triggered) 20 minutes to 2 hours (based on the cloud backup tasks in the current region) 20 minutes This function is gradually replaced
Options: N/A Default value: N/A customSan Array of strings Details: Custom Subject Alternative Name (SAN) in the server certificate of a cluster API server, which must comply with the SSL and X509 standard formats Constraints: Duplicate names are not allowed.
Click New Node on the left, enter the node name fixed-agent (which can be customized), and select Permanent Agent for Type. Specify the following node information: Number of executors: The default value is 1. Set this parameter as required.
CBR cloud server backup Master node disks, including component images, configurations, logs, and etcd data One-click backup on a web page (manually triggered) 20 minutes to 2 hours (based on the cloud backup tasks in the current region) 20 minutes This function is gradually replaced
For clusters of earlier versions (for example, earlier than v1.25.16-r10): The gRPC message size limit is 2 GiB. So, the average object size should not exceed 200 KB.
You are advised to harden the inbound rule of port 5443 for the master node security group. For details, see How Can I Configure a Security Group Rule in a Cluster? This operation will restart kube-apiserver and update the cluster access certificate (kubeconfig).
When a node fails to upgrade or restarts after the master node has been successfully upgraded, it is likely to enter the NotReady state.
It has the following advantages: Model-aware routing: Traffic can be distributed based on HTTP path matching as well as service attributes such as model names, model versions, and Low-Rank Adaptation (LoRA).
