KubeClipper

https://kubeclipper.io/docs
Admin
KubeClipper is a lightweight, easy-to-use, and production-grade Kubernetes multi-cluster lifecycle...
Tokens:23,387
Snippets:160
Trust Score:5.8
Update:1 month ago
Show doc for...
Tokens:
Raw
### Deploy 3 server and 3 agent nodes with package

Source: https://kubeclipper.io/docs/deployment-docs/ha-deploy

Deploys a cluster with 3 server and 3 agent nodes, explicitly defining the installation package.

```bash
kcctl deploy --pk-file=~/.ssh/id_rsa \
--server SERVER_IPS \
--agent AGENT_IPS \
--pkg https://oss.kubeclipper.io/release/{KC_VERSION}/kc-{GOARCH}.tar.gz
```

--------------------------------

### Deploy 3 server and 3 agent nodes without explicit package

Source: https://kubeclipper.io/docs/deployment-docs/ha-deploy

Deploys a cluster with 3 server and 3 agent nodes, defaulting to the version matching the installed kcctl.

```bash
kcctl deploy --pk-file=~/.ssh/id_rsa \
--server SERVER_IPS \
--agent AGENT_IPS
```

--------------------------------

### Deploy 3 server nodes using private key

Source: https://kubeclipper.io/docs/deployment-docs/ha-deploy

Deploys a 3-server cluster using an SSH private key and a specified installation package URL.

```bash
kcctl deploy --pk-file=~/.ssh/id_rsa \
--server SERVER_IPS \
--pkg https://oss.kubeclipper.io/release/{KC_VERSION}/kc-{GOARCH}.tar.gz
```

--------------------------------

### 查看集群状态

Source: https://kubeclipper.io/docs/getting-started

使用 kcctl get cluster 命令查看集群的当前状态。

```bash
kcctl get cluster -o yaml|grep status -A5
```

--------------------------------

### 重新创建 Kubernetes 集群

Source: https://kubeclipper.io/docs/faq/kcctl-create-cluster-error

在推送离线资源包后，再次尝试创建集群。

```bash
kcctl create cluster --name test --master 192.168.10.98 --untaint-master
```

--------------------------------

### 创建 Kubernetes 集群

Source: https://kubeclipper.io/docs/getting-started

获取节点 IP 并使用 kcctl create cluster 命令创建名为 demo 的集群。

```bash
NODE=$(kcctl get node -o yaml|grep ipv4DefaultIP:|sed 's/ipv4DefaultIP: //')

kcctl create cluster --master $NODE --name demo --untaint-master
```

--------------------------------

### 创建 Kubernetes 集群

Source: https://kubeclipper.io/docs/faq/kcctl-create-cluster-error

登录并尝试创建 Kubernetes 集群，此操作在 v1.2.1 中会触发 CNI 下载错误。

```bash
# 需要先登录
kcctl login --host http://127.0.0.1 --username admin --password Thinkbig1
# 创建集群
kcctl create cluster --name test --master 192.168.10.98 --untaint-master
```

--------------------------------

### 查看集群 Pods 状态

Source: https://kubeclipper.io/docs/faq/kcctl-create-cluster-error

检查 Kubernetes 集群中 Pods 的运行状态以确认修复结果。

```bash
kubectl get pods -A
```

--------------------------------

### 推送组件包脚本 push_component.sh

Source: https://kubeclipper.io/docs/getting-started/make-offline-package

该脚本用于下载并打包指定组件（k8s、calico、containerd、docker、metallb）的离线包，并推送到 kubeclipper。支持 amd64 和 arm64 架构。

```bash
#!/usr/bin/env bash
# usage: ./push_component.sh [component] [version] [arch]
# usage: ./push_component.sh k8s v1.23.9 amd64

set -e

if type kcctl &>/dev/null; then
  kcctl version
else
  echo "kcctl command not found, please install it first!"
fi

PKG_URL_PREFIX="https://oss.kubeclipper.io/packages"
fileList=()

name=$1
if [[ "${name}" != "k8s" ]] && [[ "${name}" != "calico" ]] && [[ "${name}" != "containerd" ]] && [[ "${name}" != "docker" ]] && [[ "${name}" != "metallb" ]]; then
  echo "only 'k8s', 'calico','containerd','docker','metallb' packages are supported. default: $name"
  exit 1
fi
version=$2
if [[ -z "${version}" ]]; then
  echo "Please specify the package version!"
  exit 1
fi
arch=$3
if [[ "${arch}" != "amd64" ]] && [[ "${arch}" != "arm64" ]]; then
  echo "only 'amd64', 'arm64' architectures are supported. default: $arch"
  exit 1
fi
``
build_dir=${name}/${version}/${arch}
pkg_name=${name}-${version}-${arch}.tar.gz
pkg_type="k8s"

packaging() {
  case $name in
  k8s)
    fileList=(
      images.tar.gz
      configs.tar.gz
      manifest.json
    )
    pkg_type="k8s"
    ;;
  calico)
    fileList=(
      images.tar.gz
      manifest.json
    )
    if [[ "$(echo -e "3.26\n${version}" | sort -V | tail -n 1)" == "${version}" ]]; then
      fileList=(
        images.tar.gz
        charts.tgz
        manifest.json
      )
    fi
    pkg_type="cni"
    ;;
  containerd|docker)
    fileList=(
      configs.tar.gz
      manifest.json
    )
    pkg_type="cri"
    ;;
  metallb)
    fileList=(
      images.tar.gz
      manifest.json
    )
    pkg_type="metallb"
    ;;
  esac

  for file in "${fileList[@]}"; do
    echo "download ${file}..."
    wget ${PKG_URL_PREFIX}/${build_dir}/${file} -P ${build_dir}
  done

  echo "packaging ${pkg_name}..."
  tar -zcvf ${pkg_name} ${build_dir}
  echo "packaging ${pkg_name} done"
  echo "push ${pkg_name} to kc..."
  kcctl resource push --pkg ${pkg_name} --type ${pkg_type}
  echo "push ${pkg_name} to kc done"
}

packaging
echo "clean up..."
rm -rf ${build_dir}
rm -rf ${pkg_name}
echo "clean up done"

```

--------------------------------

### 使用 push_component.sh 推送组件包

Source: https://kubeclipper.io/docs/getting-started/make-offline-package

示例命令展示如何使用 push_component.sh 推送 k8s、calico、containerd、docker 和 metallb 组件包到 kubeclipper。

```bash
chmod +x push_component.sh

# 推送 k8s 组件包
./push_component.sh k8s v1.33.0 amd64

# 推送 calico 组件包
./push_component.sh calico v3.26.1 amd64

# 推送 containerd 组件包
./push_component.sh containerd 1.6.4 amd64

# 推送 docker 组件包
./push_component.sh docker 20.10.20 amd64

# 推送 metallb 组件包
./push_component.sh metallb v0.13.7 amd64

```

--------------------------------

### Set KC_IMAGE_REPO_MIRROR via kc.env File

Source: https://kubeclipper.io/docs/tutorials/create-clusters

Alternative method to configure the image proxy by creating a configuration file at /etc/kc/kc.env.

```bash
# 创建 kc 目录
mkdir -pv /etc/kc

# 创建 kc.env 文件并传入环境变量信息
cat <<EOF > /etc/kc/kc.env
KC_IMAGE_REPO_MIRROR="registry.aliyuncs.com/google_containers"
EOF
```

--------------------------------

### 部署 KubeClipper AIO 环境

Source: https://kubeclipper.io/docs/faq/kcctl-create-cluster-error

使用 kcctl deploy 命令部署 AIO 环境。

```bash
# 安装 AIO 环境
kcctl deploy
```

--------------------------------

### kcctl deploy 命令语法

Source: https://kubeclipper.io/docs/getting-started/aio-env

部署 KubeClipper 时使用的命令语法，需要提供 SSH 访问凭证。

```bash
Kcctl deploy [--user <username>] [--passwd <password> | --pk-file <private key path>]
```

--------------------------------

### 推送离线资源包

Source: https://kubeclipper.io/docs/faq/kcctl-create-cluster-error

将制作好的离线资源包推送到 KubeClipper 并验证。

```bash
# 推送
kcctl resource push --pkg calico-v3.21.2-amd64.tar.gz --type cni

# 验证
kcctl resource list|grep v3.21.2
```

--------------------------------

### 查看离线组件包

Source: https://kubeclipper.io/docs/getting-started/make-offline-package

使用 kcctl resource list 命令查看已上传的离线组件包。

```bash
kcctl resource list

```

--------------------------------

### 使用 kcctl join 添加 agent 节点

Source: https://kubeclipper.io/docs/tutorials/node-management

使用 kcctl join 命令为 KubeClipper 添加 agent 节点，并可为节点指定区域。未指定区域的节点将默认归入 default 区域。

```bash
kcctl join --agent beijing:1.2.3.4 --agent shanghai:2.3.4.5
```

--------------------------------

### 配置 OIDC 身份提供商

Source: https://kubeclipper.io/docs/tutorials/access-control

在 /etc/kubeclipper-server/kubeclipepr-server.yaml 文件中的 authentication 下配置 oauthOptions 以启用 OIDC 外部用户登录。

```yaml
oauthOptions:
    identityProviders:
    - name: keycloak
      type: OIDC
      mappingMethod: auto
      provider:
        clientID: kc
        clientSecret: EErn729BB1bKawdRtnZgrqj9Bx0]mzUs
        issuer: http://172.20.163.233:7777/auth/realms/kubeclipper
        scopes:
        - openid
        - email
        redirectURL: http://{kc-console-address}/oatuh2/redirect/{IDP-Name}
```

--------------------------------

### kcctl deploy 示例

Source: https://kubeclipper.io/docs/getting-started/aio-env

使用私钥或密码进行 KubeClipper 部署的示例。

```bash
# 使用私钥
kcctl deploy --user root --pk-file /root/.ssh/id_rsa
# 使用密码
kcctl deploy --user root --passwd password
```

--------------------------------

### 挂载 NFS 存储目录

Source: https://kubeclipper.io/docs/tutorials/cluster-management

在每个 master 节点上执行此命令，将 NFS 服务器的共享目录挂载到本地路径，以便作为集群备份空间使用。

```bash
mount -t nfs { NFS_IP }:/data/kubeclipper/cluster-backups /opt/kubeclipper/cluster-backups -o proto=tcp -o nolock
```

--------------------------------

### 登录 KubeClipper

Source: https://kubeclipper.io/docs/getting-started

使用默认账号 admin 和密码 Thinkbig1 登录以获取 token，用于后续与 kc-server 交互。

```bash
kcctl login -H http://localhost -u admin -p Thinkbig1
```

--------------------------------

### 部署 KubeClipper

Source: https://kubeclipper.io/docs/getting-started

使用 kcctl deploy 命令进行安装。支持使用 SSH 密码或私钥进行身份验证。

```bash
kcctl deploy
```

```bash
kcctl deploy --user root --passwd $SSH_PASSWD
```

```bash
kcctl deploy --user root --pk-file $SSH_PRIVATE_KEY
```

--------------------------------

### Login to KubeClipper via CLI

Source: https://kubeclipper.io/docs/deployment-docs/ha-deploy

Authenticates the CLI session with the KubeClipper server using the default admin credentials.

```bash
kcctl login -H http://SERVER_IP -u admin -p Thinkbig1
```

--------------------------------

### Deploy with custom etcd ports

Source: https://kubeclipper.io/docs/deployment-docs/ha-deploy

Deploys a cluster while explicitly configuring etcd client, peer, and metric ports.

```bash
kcctl deploy --pk-file=~/.ssh/id_rsa \
--server SERVER_IPS \
--agent AGENT_IPS \
--etcd-port 12379 --etcd-peer-port 12380 --etcd-metric-port 12381
```
Tokens:
Raw