查看“︁如何搭建 Kubernetes 一站式平台”︁的源代码

----

== '''1. 什么是 Kubernetes / k8s''' ==
k8s 也就是 Kubernetes 是一个面向容器化应用的自动化管理平台, 核心职责是管理程序运行的环境, 包括 <code>部署应用/调度资源/维护副本/处理故障/对外暴露服务</code>等等...

对于初学者来说可以把它理解成一个专门管理容器集群的平台, 当你只有一个容器时也许还感觉不到它的必要性

但当你需要同时运行多个服务, 需要让服务彼此通信, 需要保证应用异常后自动恢复, 甚至需要后续扩容到多台机器时 Kubernetes 的价值就会非常明显

而本文要做的事情, 就是在一台主机上先搭建一个最小可运行的 Kubernetes 环境, 借助这个过程去理解 Kubernetes 是怎样从零开始被组织起来的

为了易读性, 在这里会将 <code>Kubernetes</code> 简称为 <code>k8s</code>

== '''2. 环境说明''' ==
运行 kubeadm 集群建议 control plane 节点至少具备 <code>2 Core CPU</code> 和 <code>2 GiB</code> 运存, 推荐购买搬瓦工的 [https://bandwagonhost.com/cart.php?a=add&pid=166 ECOMMERCE SLA] 套餐

* 操作系统以 Debian13 为例
* 主机数量 1 台
* 使用 root 用户
* 主机承担 control plane 和 workload 调度
* k8s pod网段为 <code>10.10.0.0/16</code> , service网段为 <code>192.168.0.0/16</code>
* 使用 <code>Headlamp</code> 管理k8s pods

== '''3. 前置准备''' ==

==== 3.1 加载模块并启用转发 ====
<syntaxhighlight lang="bash">
cat <<'EOF' | tee /etc/modules-load.d/k8s.conf
overlay
br_netfilter
EOF

modprobe overlay
modprobe br_netfilter

cat <<'EOF' | tee /etc/sysctl.d/k8s.conf
net.bridge.bridge-nf-call-iptables = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.ipv4.ip_forward = 1
net.ipv6.conf.all.forwarding = 1
EOF
sysctl --system
</syntaxhighlight>

==== 3.2 配置 hosts ====
<syntaxhighlight lang="bash">
tee -a /etc/hosts > /dev/null <<'EOF'
127.0.0.1 k8s-master
EOF
</syntaxhighlight>

==== 3.3 关闭 swap (可选) ====
<syntaxhighlight lang="bash">
swapoff -a
sed -ri '/\sswap\s/s/^/#/' /etc/fstab
</syntaxhighlight>

== '''4. 安装 containerd''' ==
更新软件源并安装 containerd 与 其他核心组件<syntaxhighlight lang="bash">
apt update
apt install -y containerd apt-transport-https ca-certificates curl gpg
</syntaxhighlight>生成默认配置文件并修改配置文件, 这里为了防止后续可能报错, 修改了 <code>bin_dir</code><syntaxhighlight lang="bash">mkdir -p /etc/containerd
containerd config default | tee /etc/containerd/config.toml > /dev/null
sed -i 's#bin_dir = "/usr/lib/cni"#bin_dir = "/opt/cni/bin"#' /etc/containerd/config.toml
sed -i 's/SystemdCgroup = false/SystemdCgroup = true/' /etc/containerd/config.toml</syntaxhighlight>重启并设置自启<syntaxhighlight lang="bash">
systemctl daemon-reload
systemctl restart containerd
systemctl enable containerd
systemctl status containerd
</syntaxhighlight>

== 5. 安装 kubeadm kubelet kubectl ==
<syntaxhighlight lang="bash">
mkdir -p -m 755 /etc/apt/keyrings
curl -fsSL https://pkgs.k8s.io/core:/stable:/v1.35/deb/Release.key | gpg --dearmor -o /etc/apt/keyrings/kubernetes-apt-keyring.gpg
echo 'deb [signed-by=/etc/apt/keyrings/kubernetes-apt-keyring.gpg] https://pkgs.k8s.io/core:/stable:/v1.35/deb/ /' | tee /etc/apt/sources.list.d/kubernetes.list
apt-get update
apt-get install -y kubelet kubeadm kubectl
apt-mark hold kubelet kubeadm kubectl
systemctl enable --now kubelet
</syntaxhighlight>检查<syntaxhighlight lang="bash">
kubeadm version
kubectl version --client
kubelet --version
</syntaxhighlight>

== '''6. 初始化单节点 k8s 集群''' ==
新建一个初始化配置<syntaxhighlight lang="bash">
mkdir -p /main/k8s
nano /main/k8s/kubeadm-config.yaml
</syntaxhighlight>粘贴并自行修改配置后进行保存<syntaxhighlight lang="yaml" line="1">
apiVersion: kubeadm.k8s.io/v1beta4
kind: InitConfiguration
nodeRegistration:
  name: master
  criSocket: "unix:///run/containerd/containerd.sock"
localAPIEndpoint:
  advertiseAddress: <你的公网主机IP>
  bindPort: 6443

---
apiVersion: kubeadm.k8s.io/v1beta4
kind: ClusterConfiguration
kubernetesVersion: v1.35.0
networking:
  podSubnet: 10.10.0.0/16
  serviceSubnet: 192.168.0.0/16
  dnsDomain: cluster.local

# 若开启了Swap需要解除此注释!
#---
#apiVersion: kubelet.config.k8s.io/v1beta1
#kind: KubeletConfiguration
#cgroupDriver: systemd
#failSwapOn: false
#memorySwap:
#  swapBehavior: LimitedSwap


</syntaxhighlight>开始初始化<syntaxhighlight lang="bash">
kubeadm init --config /main/k8s/kubeadm-config.yaml
</syntaxhighlight>初始化完成后配置 kubectl<syntaxhighlight lang="bash">
mkdir -p $HOME/.kube
cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
chown "$(id -u)":"$(id -g)" $HOME/.kube/config
</syntaxhighlight>检查<syntaxhighlight lang="bash">
kubectl get nodes
kubectl get pods -A
</syntaxhighlight>这时 master 大概率是 <code>NotReady</code> 状态, 目前 CNI 还没有安装, CoreDNS 也通常不会处于 <code>Running</code> 状态

== '''7. 安装 Helm''' ==
<syntaxhighlight lang="bash">curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-4
chmod 700 get_helm.sh
./get_helm.sh
helm version</syntaxhighlight>

== 8. 安装 CNI 插件 ==
在资源紧凑的主机中我更推荐使用 <code>Flannel</code> , 这里也会提供 <code>Calico</code> 与 <code>Cilium</code> 的安装, 根据应用场景自行选择其一

资源占用: <code>Flannel</code> <  <code>Calico</code> < <code>Cilium</code>

==== 8.1 Flannel ====
<syntaxhighlight lang="bash">kubectl create namespace kube-flannel
kubectl label namespace kube-flannel --overwrite pod-security.kubernetes.io/enforce=privileged
kubectl label namespace kube-flannel --overwrite pod-security.kubernetes.io/audit=privileged
kubectl label namespace kube-flannel --overwrite pod-security.kubernetes.io/warn=privileged
helm upgrade --install flannel -n kube-flannel --create-namespace --set podCidr=10.10.0.0/16 https://github.com/flannel-io/flannel/releases/download/v0.28.1/flannel.tgz</syntaxhighlight>

==== 8.2 Calico ====
<syntaxhighlight lang="bash">helm repo add projectcalico https://docs.tigera.io/calico/charts
helm repo update
kubectl create namespace tigera-operator
helm upgrade --install calico projectcalico/tigera-operator --version v3.31.4 --namespace tigera-operator --set installation.calicoNetwork.ipPools[0].cidr=10.10.0.0/16 --set installation.calicoNetwork.ipPools[0].encapsulation=VXLAN</syntaxhighlight>

==== 8.3 Cilium ====
<syntaxhighlight lang="bash">
helm repo add cilium https://helm.cilium.io/
helm repo update
helm upgrade --install cilium cilium/cilium --version 1.19.1 --namespace kube-system
</syntaxhighlight>'''检查CNI'''<syntaxhighlight lang="bash">
kubectl get nodes
kubectl get pods -A
</syntaxhighlight>节点状态从 <code>NotReady</code> 变为 <code>Ready</code> , CoreDNS 进入 <code>Running</code> 状态