kubernetes 1.18集群安裝
準備環(huán)境
本次實驗以k8s 1.8版本為例：
kubelet-1.18.0-0：在集群中的每個節(jié)點上用來啟動 Pod 和容器等
kubeadm-1.18.0-0：在集群中的每個節(jié)點上用來啟動 Pod 和容器等
kubectl-1.18.0-0：用來與集群通信的命令行工具
docker-19.03.13-ce：k8s基于docker拉取鏡像、啟動服務(wù)
準備兩臺Linux服務(wù)器（CentOS系統(tǒng)）：
控制節(jié)點：10.0.0.1，安裝docker、kubelet、kubectl和kubeadm
worker節(jié)點：10.0.0.2，安裝docker、kubelet、kubectl和kubeadm
控制節(jié)點操作
安裝docker
sudo yum install dockersudo systemctl start docker
配置kubernetes鏡像地址
cat <<EOF | sudo tee /etc/yum.repos.d/kubernetes.repo[kubernetes]name=Kubernetesbaseurl=http://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64enabled=1gpgcheck=1repo_gpgcheck=1gpgkey=http://packages.cloud.google.com/yum/doc/yum-key.gpg http://packages.cloud.google.com/yum/doc/rpm-package-key.gpgEOF
禁用SELinux，為了允許容器訪問主機文件系統(tǒng)，這么操作是為了保證后續(xù) Pod 網(wǎng)絡(luò)工作正常
setenforce 0sudo sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config
Kubernetes 1.8開始要求關(guān)閉系統(tǒng)的Swap，如果不關(guān)閉，默認配置下kubelet將無法啟動
swapoff -a安裝kubeadm、kubelet、kubectl
sudo yum install -y kubeadm-1.18.0-0 kubelet-1.18.0-0 kubectl-1.18.0-0 --disableexcludes=kubernetes查看kubeadm是否安裝成功
$ kubeadm versionkubeadm version: &version.Info{Major:"1", Minor:"8", GitVersion:"v1.8.0", GitCommit:"9befc2b8928a9426501d3bf62f72849d5cbcd5a3", GitTreeState:"clean", BuildDate:"2017-11-20T05:17:43Z", GoVersion:"go1.8.3", Compiler:"gc", Platform:"linux/amd64"}
通過systemctl 啟動kubelet
#啟動kubeletsudo systemctl start kubelet#顯示日志：Warning: kubelet.service changed on disk. Run 'systemctl daemon-reload' to reload units.# 重載所有修改過的配置文件sudo systemctl daemon-reload#開機啟動kubeletsudo systemctl enable kubelet
必須保證kubelet安裝成功，執(zhí)行下面命令不能報錯
$ kubelet version初始化kubeadm
#重置kubeadm的初始化sudo kubeadm reset# –-pod-network-cidr：用于指定Pod的網(wǎng)絡(luò)范圍，物理機ip地址，10.99必須是存在的物理機ip地址# –-service-cidr：用于指定service的網(wǎng)絡(luò)范圍，虛擬ip地址# --feature-gates：在 Kubernetes 1.18 中，用 kubeadm 來安裝 kube-dns 這一做法已經(jīng)被廢棄sudo kubeadm init --kubernetes-version=v1.18.0 --pod-network-cidr=10.244.0.0/16 --service-cidr=10.1.0.0/16
執(zhí)行初始化后，可以在初始化的日志中看到，需要在集群其他機器中執(zhí)行的命令
Your Kubernetes control-plane has initialized successfully!To start using your cluster, you need to run the following as a regular user:mkdir -p $HOME/.kubesudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/configsudo chown $(id -u):$(id -g) $HOME/.kube/configYou 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/Then you can join any number of worker nodes by running the following on each as root:kubeadm join 10.0.0.1:6443 --token wzghpa.mwvdt4ho0fn936dg \--discovery-token-ca-cert-hash sha256:aa101787f7398ac95755b1e61aa56c69cbf7205d5035184622ba8cad57abf3e1
在控制節(jié)點中配置kube相關(guān)的配置，并且export配置
mkdir -p $HOME/.kubesudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/configsudo chown $(id -u):$(id -g) $HOME/.kube/configexport KUBECONFIG=$HOME/.kube/config
執(zhí)行完上述命令后，通過kubectl version查看kubectl 是否安裝成功，能正常顯示出Client Version和Server Version等信息，即表示kubectl 集群安裝成功了，只是目前集群中只有控制節(jié)點
$ kubectl versionClient Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.0", GitCommit:"9e991415386e4cf155a24b1da15becaa390438d8", GitTreeState:"clean", BuildDate:"2020-03-25T14:58:59Z", GoVersion:"go1.13.8", Compiler:"gc", Platform:"linux/amd64"}Server Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.0", GitCommit:"9e991415386e4cf155a24b1da15becaa390438d8", GitTreeState:"clean", BuildDate:"2020-03-25T14:50:46Z", GoVersion:"go1.13.8", Compiler:"gc", Platform:"linux/amd64"}
查看集群節(jié)點
#查看集群節(jié)點$ kubectl get nodeNAME STATUS ROLES AGE VERSIONmaster NotReady master 14h v1.18.0#查看命名空間$ kubectl get nsNAME STATUS AGEdefault Active 15hkube-node-lease Active 15hkube-public Active 15hkube-system Active 15h
kubectl 安裝成功后，需要為集群部署flannel網(wǎng)絡(luò)，執(zhí)行完成以后，可以看到創(chuàng)建的一些配置
$ kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.ymlpodsecuritypolicy.policy/psp.flannel.unprivileged createdclusterrole.rbac.authorization.k8s.io/flannel createdclusterrolebinding.rbac.authorization.k8s.io/flannel createdserviceaccount/flannel createdconfigmap/kube-flannel-cfg createddaemonset.apps/kube-flannel-ds created
在命名空間kube-system中查看kube-flannel的pod是否啟動成功
$ kubectl get pod -n kube-system | grep kube-flannelkube-flannel-ds-hgkk6 0/1 Init:ImagePullBackOff 0 15m
看到kube-flannel沒有啟動成功，原因是kube-flannel的鏡像沒有下載下來，需要手動下載對應(yīng)的kube-flannel版本
$ wget https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml將kube-flannel.yml下載到服務(wù)器上后，查看yml中使用的flannel版本
$ cat kube-flannel.ymlcontainers:- name: kube-flannelimage: quay.io/coreos/flannel:v0.13.1-rc2
看到kube-flannel使用的鏡像是quay.io/coreos/flannel:v0.13.1-rc2，通過docker命令下載鏡像
$ sudo docker pull quay.io/coreos/flannel:v0.13.1-rc2$ sudo docker image ls | grep flannelquay.io/coreos/flannel v0.13.1-rc2 dee1cac4dd20 7 weeks ago 64.3MB
flannel鏡像下載下來后，kube-flannel的pod就自動啟動成功了
$ kubectl get pod -n kube-system | grep kube-flannelkube-flannel-ds-hgkk6 1/1 Running 0 6m
添加worker節(jié)點
安裝docker
sudo yum install dockersudo systemctl start docker
配置kubernetes鏡像地址
cat <<EOF | sudo tee /etc/yum.repos.d/kubernetes.repo[kubernetes]name=Kubernetesbaseurl=http://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64enabled=1gpgcheck=1repo_gpgcheck=1gpgkey=http://packages.cloud.google.com/yum/doc/yum-key.gpg http://packages.cloud.google.com/yum/doc/rpm-package-key.gpgEOF
禁用SELinux，為了允許容器訪問主機文件系統(tǒng)，這么操作是為了保證后續(xù) Pod 網(wǎng)絡(luò)工作正常
setenforce 0sudo sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config
Kubernetes 1.8開始要求關(guān)閉系統(tǒng)的Swap，如果不關(guān)閉，默認配置下kubelet將無法啟動
swapoff -a安裝kubeadm、kubelet
sudo yum install -y kubeadm-1.18.0-0 kubelet-1.18.0-0 --disableexcludes=kubernetes通過systemctl 啟動kubelet
#啟動kubeletsudo systemctl start kubelet#顯示日志：Warning: kubelet.service changed on disk. Run 'systemctl daemon-reload' to reload units.# 重載所有修改過的配置文件sudo systemctl daemon-reload#開機啟動kubeletsudo systemctl enable kubelet
接下來，需要將worker服務(wù)器，加入到k8s集群中，在worker節(jié)點服務(wù)器執(zhí)行命令
$ sudo kubeadm join 10.0.0.1:6443 --token wzghpa.mwvdt4ho0fn936dg \--discovery-token-ca-cert-hash sha256:aa101787f7398ac95755b1e61aa56c69cbf7205d5035184622ba8cad57abf3e1
如果執(zhí)行過程中，出現(xiàn)以下錯誤信息
[preflight] Some fatal errors occurred:/var/lib/kubelet is not empty[preflight] If you know what you are doing, you can skip pre-flight checks with `--skip-preflight-checks`
則需要重置kubeadm，重置后，會刪除/var/lib/kubelet下面的一些文件
$ sudo kubeadm reset重置以后，重新執(zhí)行kubeadm join命令，將worker節(jié)點加入到集群中
$ sudo kubeadm join 10.0.0.1:6443 --token wzghpa.mwvdt4ho0fn936dg \--discovery-token-ca-cert-hash sha256:aa101787f7398ac95755b1e61aa56c69cbf7205d5035184622ba8cad57abf3e1#日志輸出[preflight] Starting the kubelet service[discovery] Trying to connect to API Server "10.0.0.1:6443"[discovery] Created cluster-info discovery client, requesting info from "https://10.0.0.1:6443"[discovery] Requesting info from "https://10.0.0.1:6443" again to validate TLS against the pinned public key[discovery] Cluster info signature and contents are valid and TLS certificate validates against pinned roots, will use API Server "10.0.0.1:6443"[discovery] Successfully established connection with API Server "10.0.0.1:6443"[bootstrap] Detected server version: v1.18.4[bootstrap] The server supports the Certificates API (certificates.k8s.io/v1beta1)Node join complete:* Certificate signing request sent to master and responsereceived.* Kubelet informed of new secure connection details.Run 'kubectl get nodes' on the master to see this machine join
到此位置，就將worker節(jié)點加入了集群，最后到控制節(jié)點去執(zhí)行命令，即可看到集群中的node節(jié)點
$ kubectl get nodesNAME STATUS ROLES AGE VERSIONnode1 Ready <none> 17s v1.18.0master Ready master 19h v1.18.0
若后續(xù)還需要向集群里添加服務(wù)器，則按照worker節(jié)點安裝的流程操作即可