k8s 笔记

目录

安装

环境预准备

  1. 安装 docker:

    curl -sSL https://get.daocloud.io/docker | sh
    systemctl enable docker
    systemctl start docker
    
  2. 更改 Docker 源以及保持 Docker Cgroup Driver 和 k8s 一致:

    cat <<EOF > /etc/docker/deamon.json
    {
      "registry-mirrors": ["https://registry.cn-hangzhou.aliyuncs.com"],
      "exec-opts": ["native.cgroupdriver=systemd"]
    }
    EOF
    systemctl enable docker
    systemctl start docker
    
  3. 关闭 selinux:

    # setenforce 0
    # vim /etc/sysconfig/selinux
    SELINUX=disabled
    
  4. 关闭交换分区:

    swapoff -a
    vim /etc/fstab
    # 注释掉 swap 行
    
  5. iptables 配置:

    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
    
  6. 关闭防火墙:

    systemctl disable firewalld
    systemctl stop firewalld
    

开始安装

安装 kublet,kubeadm,kubctl,版本都是1.18.3。

  1. yum 设置源:

    cat <<EOF > /etc/yum.repos.d/kubernetes.repo
    [kubernetes]
    name=Kubernetes Respository
    baseurl=http://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
    enabled=1
    gpgcheck=0
    EOF
    
  2. 安装 kublet,kubeadm,kubctl:

    yum install -y kubelet-1.18.3 kubectl-1.18.3 kubeadm-1.18.3 --disableexcludes=kubernetes
    systemctl enable kubelet && systemctl start kubelet
    
  3. 首先在Master上安装,安装前更改配置文件:

    # kubeadm config print init-defaults > init.config.yaml
    # vim init.config.yaml
    ...
    localAPIEndpoint:
      advertiseAddress: 1.2.3.4
    ...
    # 1.2.3.4 修改为本机物理网卡ip地址,只有Master安装时需要改
    ...
    imageRepository: k8s.gcr.io
    ...
    # k8s.gcr.io 改为 registry.cn-hangzhou.aliyuncs.com/google_containers
    
  4. 查看镜像地址发现还是指向 k8s.grc.io:

    kubeadm config images list
    
  5. 可以使用之前更改过的配置文件提前拉取 kubeadm 所需镜像:

    kubeadm config images pull --config=init.config.yaml
    
  6. 开始安装Master:

    kubeadm init --config=init.config.yaml
    
  7. 完成后按照提示执行下面命令:

    mkdir -p $HOME/.kube
    sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
    sudo chown $(id -u):$(id -g) $HOME/.kube/config
    
  8. 还有一个Node加入的token命令,这里保存下来,类似:

    kubeadm join 192.168.10.8:6443 --token y3r86i.anh1x5pxt2mc680y \
        --discovery-token-ca-cert-hash sha256:9762cd94ffe6c35089d647e8b7a817226b57f21d19cafc977f41140cfaec19b9
    

    注意:如果时间过久(Node加入时间与这个token的创建时间间隔),token会失效(默认24小时),这时可以kubeadm token create --ttl 0创建永久token或者kubeadm token create创建默认24小时有效时间的token,替换上面的token之后再join即可。

  9. 开始安装Node,执行上述第1,2,3,4,5步,之后执行第8步保存下来的命令。

  10. 为了能够在Node上直接使用kubectl(不显示指定conf文件):

    cp /etc/kubernetes/kubelet.conf /etc/kubernetes/admin.conf
    
  11. 此时在Master上输入kubectl get nodes,发现 STATUS 全部是 NotReady,这是因为没有安装 cni 网络插件:

    [root@localhost ~]# kubectl get nodes
    NAME                    STATUS     ROLES    AGE     VERSION
    centos2                 NotReady   <none>   2m53s   v1.18.3
    centos3                 NotReady   <none>   117s    v1.18.3
    localhost.localdomain   NotReady   master   5h52m   v1.18.3
    

安装网络插件

在Master上安装 calico。

  1. 下载 yaml 文件:

    wget https://docs.projectcalico.org/manifests/calico.yaml
    
  2. vim 查找docker.io发现下面四个镜像:

    calico/cni:v3.21.0
    calico/pod2daemon-flexvol:v3.21.0
    calico/node:v3.21.0
    calico/kube-controllers:v3.21.0
    
  3. docker 先用之前替换好的国内源把镜像拉下来:

    docker pull calico/cni:v3.21.0 && \
    docker pull calico/pod2daemon-flexvol:v3.21.0 && \
    docker pull calico/node:v3.21.0 && \
    docker pull calico/kube-controllers:v3.21.0
    
  4. 部署 calico:

    kubectl apply -f calico.yaml
    
  5. 此时查看 pods,会看到 calico 正在初始化,等待完成后,STATUS 会全部变成 Running:

    [root@localhost ~]# kubectl get pods --all-namespaces
    NAMESPACE     NAME                                            READY   STATUS            RESTARTS   AGE
    kube-system   calico-kube-controllers-858fbfbc9-mpnjt         1/1     Running           0          96s
    kube-system   calico-node-p265c                               1/1     Running           0          97s
    kube-system   calico-node-xrsgm                               0/1     PodInitializing   0          97s
    kube-system   calico-node-z946n                               0/1     PodInitializing   0          97s
    kube-system   coredns-546565776c-cxnnf                        1/1     Running           0          7h3m
    kube-system   coredns-546565776c-f299w                        1/1     Running           0          7h3m
    kube-system   etcd-localhost.localdomain                      1/1     Running           0          7h3m
    kube-system   kube-apiserver-localhost.localdomain            1/1     Running           0          7h3m
    kube-system   kube-controller-manager-localhost.localdomain   1/1     Running           0          7h3m
    kube-system   kube-proxy-777wz                                1/1     Running           0          74m
    kube-system   kube-proxy-mj6r8                                1/1     Running           0          7h3m
    kube-system   kube-proxy-tx56l                                1/1     Running           0          73m
    kube-system   kube-scheduler-localhost.localdomain            1/1     Running           0          7h3m
    
  6. 查看 nodes,STATUS 已经全部 Ready:

    [root@localhost ~]# kubectl get nodes
    NAME                    STATUS   ROLES    AGE    VERSION
    centos2                 Ready    <none>   76m    v1.18.3
    centos3                 Ready    <none>   75m    v1.18.3
    localhost.localdomain   Ready    master   7h6m   v1.18.3
    

错误记录

  1. 安装完 Master 后:

    [root@localhost ~]# kubectl get nodes
    Unable to connect to the server: x509: certificate signed by unknown authority (possibly because of "crypto/rsa: verification error" while trying to verify candidate authority certificate "kubernetes")
    

    解决:

    [root@localhost ~]# export KUBECONFIG=/etc/kubernetes/kubelet.conf
    [root@localhost ~]# kubectl get nodes
    NAME                    STATUS   ROLES    AGE     VERSION
    localhost.localdomain   Ready    master   2m20s   v1.18.3
    
  2. 安装 calico 时出错(RBAC):

    [root@localhost ~]# kubectl apply -f calico.yaml
    Error from server (Forbidden): error when retrieving current configuration of:
    Resource: "/v1, Resource=configmaps", GroupVersionKind: "/v1, Kind=ConfigMap"
    Name: "calico-config", Namespace: "kube-system".............................
    
  3. 在 node 上执行 kubectl get pods出现:

    [root@centos3 ~]# kubectl get pods
    The connection to the server localhost:8080 was refused - did you specify the right host or port?
    

    解决:mkdir -p /root/.kube && cp /etc/kubernetes/kubelet.conf /root/.kube/config

上一篇:PHPCMS 站群调用指定站点推荐位文章


下一篇:2021SC@SDUSC-山东大学软件工程与实践-Senta(4)