Common O&M Operations on Huawei

Installing Helm 3

Mon, 01 Jan 0001 00:00:00 +0000

This section describes how to install Helm 3.

For details, see https://helm.sh/docs/intro/install/ .

Prerequisites

Ensure that the master node in the Kubernetes cluster can access the Internet.

Procedure

Run the following command to download the Helm 3 installation script.

curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3

Run the following command to modify the permission on the Helm 3 installation script.
```
chmod 700 get_helm.sh
```
Determine the Helm version to be installed based on the version mapping between Helm and Kubernetes. For details about the version mapping, see Helm Version Support Policy . Then run the following command to change the DESIRED_VERSION environment variable to the Helm version to be installed and run the installation command.

Downloading a Container Image

Mon, 01 Jan 0001 00:00:00 +0000

Downloading a Container Image Using containerd

Run the following command to download an image to a local path. In the command, image:tag indicates the image to be pulled and its tag.
```
ctr image pull <image>:<tag>
```
Run the following command to export the image to a file. In the command, image:tag indicates the image to be exported, and file indicates the name of the exported image file.
```
ctr image export <file>.tar <image>:<tag>
```

Downloading a Container Image Using Docker

Run the following command to download an image to a local path. In the command, image:tag indicates the image to be pulled.

Updating the huawei-csi-controller or huawei-csi-node Service

Mon, 01 Jan 0001 00:00:00 +0000

Perform this operation when you need to update the huawei-csi-controller or huawei-csi-node service, for example, changing the number of copies for the huawei-csi-controller service.

Procedure

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Go to the /helm/esdk directory and run the following command to obtain the original service configuration file. helm-huawei-csi indicates the Helm chart name specified during the installation of the earlier version, and huawei-csi indicates the Helm chart namespace specified during the installation of the earlier version. For details about the component package path, see Table 1 .

Modifying the Log Output Mode

Mon, 01 Jan 0001 00:00:00 +0000

huawei-csi supports two log output modes: file and console. file indicates that logs are output to the fixed directory (/var/log/huawei), and console indicates that logs are output to the standard directory of the container. You can set the log output mode as required. The default mode is file.

Procedure

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.

Enabling the ReadWriteOncePod Feature Gate

Mon, 01 Jan 0001 00:00:00 +0000

The ReadWriteOnce access mode is the fourth access mode introduced by Kubernetes v1.22 for PVs and PVCs. If you create a Pod using a PVC in ReadWriteOncePod access mode, Kubernetes ensures that the Pod is the only Pod in the cluster that can read or write the PVC.

If the Kubernetes version is earlier than v1.29, you need to enable the ReadWriteOncePod feature in feature-gates of kube-apiserver, kube-scheduler, and kubelet.

Configuring Access to the Kubernetes Cluster as a Non-root User

Mon, 01 Jan 0001 00:00:00 +0000

Procedure

Copy the authentication file of the Kubernetes cluster and modify /etc/kubernetes/admin.conf to be the actual authentication file.
```
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
```
Change the user and user group of the authentication file.
```
sudo chown $(id -u):$(id -g) $HOME/.kube/config
```
Configure the KUBECONFIG environment variable of the current user. The following uses Ubuntu 20.04 as an example.
```
echo "export KUBECONFIG=$HOME/.kube/config" >> ~/.bashrc
source ~/.bashrc
```

Configuring Storage Topology Awareness

Mon, 01 Jan 0001 00:00:00 +0000

In the Kubernetes cluster, resources can be scheduled and provisioned based on the topology labels of nodes and the topology capabilities supported by storage backends.

Prerequisites

You need to configure topology labels on worker nodes in the cluster. The method is as follows:

Use a remote access tool, such as PuTTY, to log in to any master node in the Kubernetes cluster through the management IP address.
Run the following command to view information about worker nodes in the current cluster.