ubuntu 20.04下用 kubeadm 安装 kubenetes
通过 kubeadm 在 ubuntu 上安装 kubenetes
参考 Kubernetes 官方文档:
前期准备
关闭防火墙
systemctl disable firewalld && systemctl stop firewalld
安装docker和bridge-utils
要求节点上安装有 docker (或者其他container runtime）和 bridge-utils (用来操作linux bridge).
查看 docker 版本：
$ docker --version
Docker version 20.10.21, build baeda1f
bridge-utils可以通过apt安装：
sudo apt-get install bridge-utils
设置iptables
要确保 br_netfilter 模块已经加载,可以通过运行 lsmod | grep br_netfilter来完成。
$ lsmod | grep br_netfilter
br_netfilter 32768 0
bridge 307200 1 br_netfilter
如需要明确加载，请调用 sudo modprobe br_netfilter。
为了让作为Linux节点的iptables能看到桥接流量，应该确保 net.bridge.bridge-nf-call-iptables 在 sysctl 配置中被设置为1，执行命令：
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
禁用虚拟内存swap
执行 free -m 命令检测:
$ free -m
total used free shared buff/cache available
Mem: 15896 1665 11376 20 2854 13819
Swap: 0 0 0
如果Swap这一行不是0，则说明虚拟内存swap被开启了，需要关闭。
需要做两个事情：
-
操作系统安装时就不要设置swap分区，如果有，删除该swap分区
-
即使没有swap分区，也会开启swap，需要通过
sudo vi /etc/fstab找到swap 这一行：# 在swap分区这行前加 # 禁用掉swap /swapfile none swap sw 0 0重启之后再用
free -m命令检测。
设置docker的cgroup driver
docker 默认的 cgroup driver 是 cgroupfs，可以通过 docker info 命令查看：
$ docker info | grep "Cgroup Driver"
Cgroup Driver: cgroupfs
而 kubernetes 在v1.22版本之后，如果用户没有在 KubeletConfiguration 下设置 cgroupDriver 字段，则 kubeadm 将默认为 systemd。
需要修改 docker 的 cgroup driver 为 systemd, 方式为打开 docker 的配置文件（如果不存在则创建）
sudo vi /etc/docker/daemon.json
增加内容：
{
"exec-opts": ["native.cgroupdriver=systemd"]
}
修改完成后重启 docker：
systemctl restart docker
# 重启后检查一下
docker info | grep "Cgroup Driver"
否则，在安装过程中，由于 cgroup driver 的不一致，kubeadm init 命令会因为 kubelet 无法启动而超时失败，报错为:
[wait-control-plane] Waiting for the kubelet to boot up the control plane as static Pods from directory "/etc/kubernetes/manifests". This can take up to 4m0s
[kubelet-check] Initial timeout of 40s passed.
Unfortunately, an error has occurred:
timed out waiting for the condition
This error is likely caused by:
- The kubelet is not running
- The kubelet is unhealthy due to a misconfiguration of the node in some way (required cgroups disabled)
If you are on a systemd-powered system, you can try to troubleshoot the error with the following commands:
- 'systemctl status kubelet'
- 'journalctl -xeu kubelet'
执行 systemctl status kubelet 会发现 kubelet 因为报错而退出，执行 journalctl -xeu kubelet 会发现有如下的错误信息：
Dec 26 22:31:21 skyserver2 kubelet[132861]: I1226 22:31:21.438523 132861 docker_service.go:264] "Docker Info" dockerInfo=&{ID:AEON:SBVF:43UK:WASV:YIQK:QGGA:7RU3:IIDK:DV7M:6QLH:5ICJ:KT6R Containers:2 ContainersRunning:0 ContainersPaused:>
Dec 26 22:31:21 skyserver2 kubelet[132861]: E1226 22:31:21.438616 132861 server.go:302] "Failed to run kubelet" err="failed to run Kubelet: misconfiguration: kubelet cgroup driver: \"systemd\" is different from docker cgroup driver: \"c>
Dec 26 22:31:21 skyserver2 systemd[1]: kubelet.service: Main process exited, code=exited, status=1/FAILURE
-- Subject: Unit process exited
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- An ExecStart= process belonging to unit kubelet.service has exited.
--
-- The process' exit code is 'exited' and its exit status is 1.
参考：
- https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/configure-cgroup-driver/
- https://blog.51cto.com/riverxyz/2537914
安装kubeadm
切记
想办法搞定全局翻墙，不然kubeadm安装是非常麻烦的。按照官方文档的指示，执行如下命令：
sudo -i
apt-get update
apt-get install -y apt-transport-https ca-certificates curl
curl -fsSLo /usr/share/keyrings/kubernetes-archive-keyring.gpg https://packages.cloud.google.com/apt/doc/apt-key.gpg
echo "deb [signed-by=/usr/share/keyrings/kubernetes-archive-keyring.gpg] https://apt.kubernetes.io/ kubernetes-xenial main" | sudo tee /etc/apt/sources.list.d/kubernetes.list
apt-get update
apt-get install -y kubelet kubeadm kubectl
这会安装最新版本的kubernetes:
......
Setting up conntrack (1:1.4.6-2build2) ...
Setting up kubectl (1.25.4-00) ...
Setting up ebtables (2.0.11-4build2) ...
Setting up socat (1.7.4.1-3ubuntu4) ...
Setting up cri-tools (1.25.0-00) ...
Setting up kubernetes-cni (1.1.1-00) ...
Setting up kubelet (1.25.4-00) ...
Created symlink /etc/systemd/system/multi-user.target.wants/kubelet.service → /lib/systemd/system/kubelet.service.
Setting up kubeadm (1.25.4-00) ...
Processing triggers for man-db (2.10.2-1) ...
Processing triggers for doc-base (0.11.1) ...
Processing 1 added doc-base file...
# 查看版本
$ kubeadm version
kubeadm version: &version.Info{Major:"1", Minor:"25", GitVersion:"v1.25.4", GitCommit:"872a965c6c6526caa949f0c6ac028ef7aff3fb78", GitTreeState:"clean", BuildDate:"2022-11-09T13:35:06Z", GoVersion:"go1.19.3", Compiler:"gc", Platform:"linux/amd64"}
$ kubelet --version
Kubernetes v1.25.4
$ kubectl version
Client Version: version.Info{Major:"1", Minor:"25", GitVersion:"v1.25.4", GitCommit:"872a965c6c6526caa949f0c6ac028ef7aff3fb78", GitTreeState:"clean", BuildDate:"2022-11-09T13:36:36Z", GoVersion:"go1.19.3", Compiler:"gc", Platform:"linux/amd64"}
Kustomize Version: v4.5.7
The connection to the server localhost:8080 was refused - did you specify the right host or port?
如果希望安装特定版本：
apt-get install kubelet=1.23.5-00 kubeadm=1.23.5-00 kubectl=1.23.5-00
apt-get install kubelet=1.23.14-00 kubeadm=1.23.14-00 kubectl=1.23.14-00
apt-get install kubelet=1.24.8-00 kubeadm=1.24.8-00 kubectl=1.24.8-00
具体有哪些可用的版本，可以看这里：
https://packages.cloud.google.com/apt/dists/kubernetes-xenial/main/binary-amd64/Packages
由于 kubernetes 1.25 之后默认使用
安装k8s
同样切记
想办法搞定全局翻墙。sudo kubeadm init --pod-network-cidr=10.244.0.0/16 -v=9
sudo kubeadm init --pod-network-cidr=10.244.0.0/16 --apiserver-advertise-address=192.168.100.40 -v=9
注意后面为了使用 CNI network 和 Flannel，我们在这里设置了 --pod-network-cidr=10.244.0.0/16，如果不加这个设置，Flannel 会一直报错。如果机器上有多个网卡，可以用 --apiserver-advertise-address 指定要使用的IP地址。
如果遇到报错:
[preflight] Some fatal errors occurred:
[ERROR CRI]: container runtime is not running: output: E1125 11:16:01.799551 14661 remote_runtime.go:948] "Status from runtime service failed" err="rpc error: code = Unimplemented desc = unknown service runtime.v1alpha2.RuntimeService"
time="2022-11-25T11:16:01+08:00" level=fatal msg="getting status of runtime: rpc error: code = Unimplemented desc = unknown service runtime.v1alpha2.RuntimeService"
, error: exit status 1
则可以执行下列命令之后重新尝试 kubeadm init：
$ rm -rf /etc/containerd/config.toml
$ systemctl restart containerd.service
kubeadm init 输出如下：
......
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/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
Alternatively, if you are the root user, you can run:
export KUBECONFIG=/etc/kubernetes/admin.conf
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/
Then you can join any number of worker nodes by running the following on each as root:
kubeadm join 192.168.100.40:6443 --token uq5nqn.bppygpcqty6icec4 \
--discovery-token-ca-cert-hash sha256:51c13871cd25b122f3a743040327b98b1c19466d01e1804aa2547c047b83632b
为了使用普通用户，按照上面的提示执行：
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
安装完成后，node处于NotReady状态：
$ kubectl get node
NAME STATUS ROLES AGE VERSION
skyserver NotReady control-plane,master 3m7s v1.23.5
kubectl describe 可以看到是因为没有安装 network plugin
$ kubectl describe node skyserver
Name: skyserver
Roles: control-plane,master
......
Ready False Thu, 24 Mar 2022 13:57:21 +0000 Thu, 24 Mar 2022 13:57:06 +0000 KubeletNotReady container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:docker: network plugin is not ready: cni config uninitialized
安装flannel：
kubectl apply -f https://raw.githubusercontent.com/flannel-io/flannel/master/Documentation/kube-flannel.yml
备注：有时会遇到 raw.githubusercontent.com 这个域名被污染，解析为 127.0.0.1，导致无法访问。解决方法是访问 https://ipaddress.com/website/raw.githubusercontent.com 然后查看可用的IP地址，找一个速度最快的，在
/etc/hosts文件中加入一行记录即可，如185.199.111.133 raw.githubusercontent.com。
稍等就可以看到 node 的状态变为 Ready了：
$ kubectl get node
NAME STATUS ROLES AGE VERSION
skyserver Ready control-plane,master 4m52s v1.23.5
最后，如果是测试用的单节点，为了让负载可以跑在k8s的master节点上，执行下列命令去除master的污点：
kubectl taint nodes --all node-role.kubernetes.io/master-
可以通过 kubectl describe node skyserver 对比去除污点前后 node 信息中的 Taints 部分，去除污点前：
Taints: node.kubernetes.io/not-ready:NoExecute
node-role.kubernetes.io/master:NoSchedule
node.kubernetes.io/not-ready:NoSchedule
去除污点后：
Taints: <none>
常见问题
有时会遇到 coredns pod无法创建的情况:
$ k get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
kube-system coredns-64897985d-9z82d 0/1 ContainerCreating 0 82s
kube-system coredns-64897985d-wkzc7 0/1 ContainerCreating 0 82s
问题发生在 flannel 上：
$ k describe pods -n kube-system coredns-64897985d-9z82d
......
Warning FailedCreatePodSandBox 100s kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to set up sandbox container "675b91ac9d25f0385d3794847f47c94deac2cb712399c21da59cf90e7cccb246" network for pod "coredns-64897985d-9z82d": networkPlugin cni failed to set up pod "coredns-64897985d-9z82d_kube-system" network: open /run/flannel/subnet.env: no such file or directory
Normal SandboxChanged 97s (x12 over 108s) kubelet Pod sandbox changed, it will be killed and re-created.
Warning FailedCreatePodSandBox 96s (x4 over 99s) kubelet (combined from similar events): Failed to create pod sandbox: rpc error: code = Unknown desc = failed to set up sandbox container "b46dcd8abb9ab0787fdb2ab9f33ebf052c2dd1ad091c006974a3db7716904196" network for pod "coredns-64897985d-9z82d": networkPlugin cni failed to set up pod "coredns-64897985d-9z82d_kube-system" network: open /run/flannel/subnet.env: no such file or directory
解决的方式就是重新执行:
kubectl apply -f https://raw.githubusercontent.com/flannel-io/flannel/master/Documentation/kube-flannel.yml
备注：这个问题只遇到过一次。
失败重来
如果遇到安装失败，需要重新开始，或者想铲掉现有的安装，则可以：
- 运行
kubeadm reset - 删除
.kube目录 - 再次执行
kubeadm init
如果网络设置有改动，则需要彻底的重置网络。具体见下一章。
将节点加入到集群
如果有多个kubenetes节点（即多台机器），则需要将其他节点加入到集群中。具体见下一章。