Install the Kubernetes

The Kubernetes version should be v1.24.0+

Table of Content


Before you install the Computing Provider, you need to know there are some resources required:

  • Possess a public IP

  • Have a domain name (*

  • Have an SSL certificate

  • Go version must 1.19+, you can refer here:

wget -c -O - | sudo tar -xz -C /usr/local

echo "export PATH=$PATH:/usr/local/go/bin" >> ~/.bashrc && source ~/.bashrc

Install the Kubernetes

The Kubernetes version should be v1.24.0+

Install Container Runtime Environment

If you plan to run a Kubernetes cluster, you need to install a container runtime into each node in the cluster so that Pods can run there, refer to here. And you just need to choose one option to install the Container Runtime Environment

Option 1: Install the Docker and cri-dockerd (Recommended)

To install the Docker Container Runtime and the cri-dockerd, follow the steps below:

  • Install the Docker:

    • Please refer to the official documentation from here.

  • Install cri-dockerd:

    • cri-dockerd is a CRI (Container Runtime Interface) implementation for Docker. You can install it refer to here.

Option 2: Install the Containerd

Containerd is an industry-standard container runtime that can be used as an alternative to Docker. To install containerd on your system, follow the instructions on getting started with containerd.

Optional-Setup a docker registry server

If you are using the docker and you have only one node, the step can be skipped.

If you have deployed a Kubernetes cluster with multiple nodes, it is recommended to set up a private Docker Registry to allow other nodes to quickly pull images within the intranet.

  • Create a directory /docker_repo on your docker server. It will be mounted on the registry container as persistent storage for our docker registry.

sudo mkdir /docker_repo
sudo chmod -R 777 /docker_repo
  • Launch the docker registry container:

sudo docker run --detach \
  --restart=always \
  --name registry \
  --volume /docker_repo:/docker_repo \
  --publish 5000:5000 \
  • Add the registry server to the node

    • If you have installed the Docker and cri-dockerd(Option 1), you can update every node's configuration:

    sudo vi /etc/docker/daemon.json
    ## Add the following config
    "insecure-registries": ["<Your_registry_server_IP>:5000"]

    Then restart the docker service

    sudo systemctl restart docker
    • If you have installed the containerd(Option 2), you can update every node's configuration:

      endpoint = ["http://<Your_registry_server_IP>:5000"]

      insecure_skip_verify = true                                                               

Then restart containerd service

sudo systemctl restart containerd

<Your_registry_server_IP>: the intranet IP address of your registry server.

Finally, you can check the installation by the command:

docker system info

Create a Kubernetes Cluster

To create a Kubernetes cluster, you can use a container management tool like kubeadm. The below steps can be followed:

Install the Network Plugin

Calico is an open-source networking and network security solution for containers, virtual machines, and native host-based workloads. Calico supports a broad range of platforms including Kubernetes, OpenShift, Mirantis Kubernetes Engine (MKE), OpenStack, and bare metal services.

To install Calico, you can follow the below steps, more information can be found here.

step 1: Install the Tigera Calico operator and custom resource definitions

kubectl create -f

step 2: Install Calico by creating the necessary custom resource

kubectl create -f

step 3: Confirm that all of the pods are running with the following command

watch kubectl get pods -n calico-system

step 4: Remove the taints on the control plane so that you can schedule pods on it.

kubectl taint nodes --all
kubectl taint nodes --all

If you have installed it correctly, you can see the result shown in the figure by the command kubectl get po -A


  • If you are a single-host Kubernetes cluster, remember to remove the taint mark, otherwise, the task can not be scheduled to it.

kubectl taint node ${nodeName}

Install the NVIDIA Plugin

If your computing provider wants to provide a GPU resource, the NVIDIA Plugin should be installed, please follow the steps:

Recommend NVIDIA Linux drivers version should be 470.xx+

If you have installed it correctly, you can see the result shown in the figure by the command kubectl get po -n kube-system

Install the Ingress-nginx Controller

The ingress-nginx is an ingress controller for Kubernetes using NGINX as a reverse proxy and load balancer. You can run the following command to install it:

kubectl apply -f

If you have installed it correctly, you can see the result shown in the figure by the command:

  • Run kubectl get po -n ingress-nginx

  • Run kubectl get svc -n ingress-nginx

Last updated