The environment

  • Centos 7.6
  • K8s 1.13.4
  • 3 machines, 1 master, 2 workers

The preparatory work

Close the swap

Run the swapoff command to temporarily close the swap. Note The swap partition will become invalid after a restart. To permanently disable the swap partition, you can comment out a line in the /etc/fstab file

As for why shut down here is a description: https://github.com/kubernetes/kubernetes/issues/53533, https://www.zhihu.com/question/374752553 has also said that affect performance

Disable the firewall and Selinux

According to the document: https://kubernetes.io/zh/docs/setup/production-environment/tools/kubeadm/install-kubeadm/

Setenforce 0 sed -i 's/^ SELinux =enforcing$/ SELinux =permissive/' /etc/selinux/configCopy the code

Open ports

Allows iptables to check bridge traffic

cat <<EOF | sudo tee /etc/modules-load.d/k8s.conf
br_netfilter
EOF

cat <<EOF | sudo tee /etc/sysctl.d/k8s.conf
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
EOF
sudo sysctl --system

Copy the code

Install Docker (all nodes)

The installation

Y y y y y y y y y y y y y y y y y y y y y y y y y y y y y http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo # to check what are the docker version yum list docker - ce - showduplicates | Sort -r # yum makecache fast && yum install -y docker-ce-18.09.8-3.el7 docker-ce-cli-18.09.8-3.el7 Containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7: containerd.io-1.2.0-3.el7 docker.serviceCopy the code

Modify the default Docker storage location

Systemctl stop docker or service docker stop # then move the entire /var/lib/docker directory to the destination path: Mv /var/lib/docker /home/data/docker ln -s /home/data/docker /var/lib/docker #reload Configuration file systemctl daemon-reload // You can also modify the configuration file by way of vim /etc/docker/daemon.json {"registry-mirrors": ["http://7e61f7f9.m.daocloud.io"],"graph": "/new-path/docker"}Copy the code

Ali Cloud image acceleration

# # visit: https://cr.console.aliyun.com/cn-beijing/instances/mirrors to find acceleration methods, such as:  sudo mkdir -p /etc/docker sudo tee /etc/docker/daemon.json <<-'EOF' { "registry-mirrors": ["https://se35r65b.mirror.aliyuncs.com"] } EOF sudo systemctl daemon-reload sudo systemctl restart dockerCopy the code

Install kubeadm, kubelet and kubectl(master and worker installed)

Add to create the/etc/yum yum warehouse. Repos. D/kubernetes. Repo, file content as follows

[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
enabled=1
gpgcheck=1
repo_gpgcheck=1
gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg https://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
Copy the code

Install kubelet kubectl and kubeadm

Yum install -y kubelet-1.13.4 kubeadm-1.13.4 kubectl-1.13.4 kubernetes-cni-0.6.0 systemctl enable --now kubeletCopy the code

Manually pick image from Aliyun

Run kubeadm config images pull to check the connection to gcr. IO. If the pull succeeds, go to the next step. If it fails, grc.io cannot be accessed. In this case, you need to manually pull the image. You can execute the following script to pull the image from Ali Cloud

#! /bin/bash images=(kube-apiserver:v1.13.4 kube-controller-manager:v1.13.4 kube-scheduler:v1.13.4 kube-proxy:v1.13.4 Etcd :3.2.24 coreDNS :1.2.6) for imageName in ${images[@]}; do docker pull registry.cn-hangzhou.aliyuncs.com/google_containers/$imageName docker tag registry.cn-hangzhou.aliyuncs.com/google_containers/$imageName k8s.gcr.io/$imageName doneCopy the code

Initialize (master)

Remember to add pod-network-CIDR because the following network components use flannel

Kubeadm init - pod - network - cidr = 10.244.0.0/16 - image - repository registry.aliyuncs.com/google_containersCopy the code

Installation success message

Your Kubernetes master has initialized successfully! To start using your cluster, you need to run the following as a regular user: mkdir -p $HOME/.kube sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config sudo chown $(id -u):$(id -g) $HOME/.kube/config You should now deploy a pod network to the cluster. Run "kubectl apply -f [podnetwork].yaml" with one  of the options listed at: https://kubernetes.io/docs/concepts/cluster-administration/addons/ You can now join any number of machines by running the following on each node as root: Kubeadm join 10.22.9.162:6443 --token e225cp.14g848dy4vPOas75 --discovery-token-ca-cert-hash sha256:aaf9910fb2b94e8c2bc2aea0b2a08538796d8322331561ef1094bebe8a7a790fCopy the code

This is the first time to use the Kubernetes cluster configuration command

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
Copy the code

The reason for these configuration commands is that the Kubernetes cluster requires encrypted access by default. Kubectl will use the authorization information in this directory to access Kubernetes cluster by default. Kubectl will use the authorization information in this directory to access Kubernetes cluster. Otherwise, we would need to tell Kubectl the location of the security profile each time through the export KUBECONFIG environment variable.

The master node generates the mode for other nodes to join

kubeadm token create --print-join-command
Copy the code

Deploy the Flannel network component

kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/a70459be0084506e4ec919aa1c114638878db11b/Documentation/kube-flannel.yml
Copy the code

Check the status

# use kubectl the get command to view the current state of a Node only the kubectl get # nodes with kubectl go to view the details of this Node (the Node) object (Event) kubectl, state, and events Describe node master # re-check Pod status via kubectl get Kubectl get Pods -n kube-system # Journalctl -l -u kubeletCopy the code

Master Node Configuration

  • Delete the default stain on the master node

Taint: I don’t know. By default, the cluster does not schedule pods on the master node. If you prefer to schedule pods on the master node, you can do the following:

# check the stain kubectl the describe the node master | grep -i taints # remove stain kubectl taint nodes master node - role. Kubernetes. IO/master -Copy the code

Joining a cluster (worker)

Join the cluster using the information previously initialized by the master

Kubeadm join 10.22.9.162:6443 --token 43t2na.80oiehldy76rw6lz --discovery-token-ca-cert-hash sha256:67fd28cb6fd03242eda63c7a395096aba1a6784f7234a9b6269ff0941e9070e3Copy the code

Check the cluster status on the master

kubectl get nodes
Copy the code

Installing the Dashboard UI (Master)

Get the configuration file

Wget HTTP: / / https://raw.githubusercontent.com/kubernetes/dashboard/v1.10.1/src/deploy/recommended/kubernetes-dashboard.yamlCopy the code

Obtaining a Mirror Manually

Docker pull anjia0532/ Google-containers. kubernetes-dashboards-AMD64 :v1.10.0 Docker tag Anjia0532 / Google - containers. Kubernetes dashboard - amd64: v1.10.0 k8s. GCR. IO/kubernetes - dashboard - amd64: v1.10.0 docker rmi Anjia0532 / Google - containers. Kubernetes dashboard - amd64: v1.10.0Copy the code

Modifying the Configuration File (Ports)

# ------------------- Dashboard Service ------------------- #
kind: Service
apiVersion: v1
metadata:
  labels:
    k8s-app: kubernetes-dashboard
  name: kubernetes-dashboard
  namespace: kube-system
spec:
  type: NodePort
  ports:
    - port: 443
      targetPort: 8443
      nodePort: 30001
  selector:
    k8s-app: kubernetes-dashboard
Copy the code

Run and view the status

Yaml # kubectl get Pods -n kubernetes-dashboard kubectl get pods,svc -n kubernetes-dashboardCopy the code

The login

Kubectl create ServiceAccount dashboard-admin -n kube-system kubectl create ClusterRoleBinding dashboard-admin -- ClusterRole =cluster-admin -- ServiceAccount =kube-system: dashdash-admin ## Obtain token kubectl describe secrets -n kube-system $(kubectl -n kube-system get secret | grep dashboard-admin | awk '{print $1}')Copy the code

Clear or uninstall K8s completely

This a gist for quick uninstall kubernetes If the cluster is node, First delete it from master

Kubectl drain <node name> -- delete-local-data -- force -- ignore-daemonsets kubectl delete node <node name>Copy the code

Then remove kubeadm completely

kubeadm reset 
# on debian base 
sudo apt-get purge kubeadm kubectl kubelet kubernetes-cni kube* 
#on centos base
sudo yum remove kubeadm kubectl kubelet kubernetes-cni kube*
# on debian base
sudo apt-get autoremove
#on centos base
sudo yum autoremove
 
sudo rm -rf ~/.kube
Copy the code

Reference:

  • www.yinxiang.com/everhub/not…

  • Blog.51cto.com/3241766/240…

  • Juejin. Cn/post / 684490…