kubekey

English | 中文

👋 Welcome to KubeKey!

KubeKey is an open-source lightweight task flow execution tool. It provides a flexible and fast way to install Kubernetes.

KubeKey has passed the CNCF Kubernetes Conformance Certification

Comparison of new features in 3.x

1. Expanded from Kubernetes lifecycle management tool to task execution tool (flow design refers to Ansible)
2. Supports multiple ways to manage task templates: git, local, etc.
3. Supports multiple node connection methods, including: local, ssh, kubernetes, prometheus.
4. Supports cloud-native automated batch task management
5. Advanced features: UI page (not yet open)

Install kubekey

Install in Kubernetes

Install kubekey via helm.

helm upgrade --install --create-namespace -n kubekey-system kubekey config/kubekey

Binary

Get the corresponding binary files from the release page.

Download Binary with UI

UI only support after v4.0.0

Prerequisite: hack/downloadKubekey.sh downloads the web-installer bundle unless SKIP_WEB_INSTALLER is set to true.

export SKIP_WEB_INSTALLER=false
curl -sfL https://get-kk.kubesphere.io | sh -
# run with UI
kk web --schema-path schema --ui-path dist

Build an offline package with config.yaml

After hack/downloadKubekey.sh finishes, you may have package.sh in the current directory (depending on version and options). Put your config.yaml there and run ./package.sh config.yaml to produce the offline bundle. To generate or fill in config.yaml—including image lists and related fields—use the online tool KubeSphere Images.

Deploy Kubernetes

• Supported deployment environments: Linux distributions

  • almaLinux: 9.0 (not fully tested)
  • centOS: 8
  • debian: 10, 11
  • kylin: V10SP3 (not fully tested)
  • ubuntu: 18.04, 20.04, 22.04, 24.04.

• Supported Kubernetes versions: v1.23.x ~ v1.34.x

Requirements

• One or more computers running Linux operating systems compatible with deb/rpm; for example: Ubuntu or CentOS.
• Each machine should have more than 2 GB of memory; applications will be limited if memory is insufficient.
• Control plane nodes should have at least 2 CPUs.
• Full network connectivity among all machines in the cluster. You can use public or private networks.

Define node information

kubekey uses the inventory resource to define node connection information.
You can use kk create inventory to get the default inventory.yaml resource. The default inventory.yaml configuration is as follows:

apiVersion: kubekey.kubesphere.io/v1
kind: Inventory
metadata:
  name: default
spec:
  hosts: # your can set all nodes here. or set nodes on special groups.
#    node1:
#      connector:
#        type: ssh
#        host: node1
#        port: 22
#        user: root
#        password: 123456
  groups:
    # all kubernetes nodes.
    k8s_cluster:
      groups:
        - kube_control_plane
        - kube_worker
    # control_plane nodes
    kube_control_plane:
      hosts:
        - localhost
    # worker nodes
    kube_worker:
      hosts:
        - localhost
    # etcd nodes when etcd_deployment_type is external
    etcd:
      hosts:
        - localhost
#    image_registry:
#      hosts:
#        - localhost
    # nfs nodes for registry storage. and kubernetes nfs storage
#    nfs:
#      hosts:
#        - localhost

The inventory contains the following built-in groups:

1. k8s_cluster: Kubernetes cluster. Contains two subgroups: kube_control_plane, kube_worker
2. kube_control_plane: control_plane node group in the Kubernetes cluster
3. kube_worker: worker node group in the Kubernetes cluster.
4. etcd: node group for installing etcd cluster.
5. image_registry: node group for installing image registry (including harbor, registry)
6. nfs: node group for installing nfs.

Define key configuration information

kubekey uses the config resource to define node connection information.
You can use kk create config --with-kubernetes v1.33.1 to get the default inventory.yaml resource. The default config.yaml configuration is as follows:

Default config configurations are provided as references for different Kubernetes versions:

• Config for installing Kubernetes v1.23.x
• Config for installing Kubernetes v1.24.x
• Config for installing Kubernetes v1.25.x
• Config for installing Kubernetes v1.26.x
• Config for installing Kubernetes v1.27.x
• Config for installing Kubernetes v1.28.x
• Config for installing Kubernetes v1.29.x
• Config for installing Kubernetes v1.30.x
• Config for installing Kubernetes v1.31.x
• Config for installing Kubernetes v1.32.x
• Config for installing Kubernetes v1.33.x
• Config for installing Kubernetes v1.34.x

Install cluster

You can create a cluster in two ways: use the Web UI, or use the command line. The prerequisites and the meaning of inventory / config are described in the sections above.

Method one: Web (UI)

Requires KubeKey v4.0.0 or newer with the web installer bundle. Install or download that build (see Download Binary with UI under Install kubekey), then start the UI:

kk web --schema-path schema --ui-path dist

In the browser, edit inventory and configuration, then follow the UI flow to run the cluster creation playbook (equivalent to playbooks/create_cluster.yaml).

Method two: Command line

Prepare inventory.yaml and config.yaml, then run the built-in kk create cluster subcommand; it runs playbooks/create_cluster.yaml for you, so you do not pass the playbook path explicitly.

kk create cluster -i inventory.yaml -c config.yaml

If -i inventory.yaml is not provided, the default inventory.yaml is used. Kubernetes will only be installed on the executing machine. If -c config.yaml is not provided, the default config.yaml is used. Installs Kubernetes version v1.34.1.

Other useful flags:

• --workdir: KubeKey working directory (default: <current-dir>/kubekey).
• --with-kubernetes: Kubernetes version (for example v1.33.1) when it is not set in config.yaml.
• -a / --artifact: path to an offline KubeKey artifact package (.tgz); also turns off online fetching when set.
• --set: override config fields, for example --set download.fetch=false or nested keys supported by KubeKey.
• -n / --namespace: namespace for local runtime resources tied to the playbook.

Offline install example:

kk create cluster -i inventory.yaml -c config.yaml --workdir ./kubekey --artifact /path/to/kubekey-artifact.tgz

Documentation

Custom Playbook
Kubernetes Playbook