其他分享
首页 > 其他分享> > k8s 笔记

k8s 笔记

作者：互联网

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

标签：k8s,system,笔记,docker,kube,config,calico,localhost
来源： https://www.cnblogs.com/coodyz/p/15860743.html