Kubernetes集群部署
本次通过kubeadm安装 1.29.2 版本
前言
k8s启动的流程 linux > docker > cri-docker > kberclt > AS/CM/SCH - AS apiserver - CM controller-manager - SCH scheduler
通过kubeadm的安装方式 组件以容器化的方式运行 优势 - 简单 - 自愈 缺点 - 集群部署掩盖一些细节不便于理解 二进制安装 组件以系统进程的方式运行 优势 - 能够更灵活的配置集群 缺点 - 难理解
一、准备工作
下载rockylinux9镜像
官方地址
https://rockylinux.org/zh-CN/download
根据系统架构选择镜像
阿里云镜像站地址
https://mirrors.aliyun.com/rockylinux/9/isos/x86_64/?spm=a2c6h.25603864.0.0.29696621VzJej5
二、系统配置
系统核心数：4c 内存：4G 磁盘：100G 一主两从架构
设备掩码255.255.255.0
设备网关192.168.68.1
设备IP master 192.168.68.160 node1 192.168.68.161 node2 192.168.68.162
固定所有设备的IP地址，配置仅需要修改ipv4块
# 网卡配置 # cat /etc/NetworkManager/system-connections/enp6s18.nmconnection [ipv4] method=manual address=192.168.68.160/24,192.168.68.1 dns=223.5.5.5;119.29.29.29
配置系统镜像源
https://developer.aliyun.com/mirror/rockylinux?spm=a2c6h.13651102.0.0.47731b11u4haJL # 阿里云镜像源
sed -e 's|^mirrorlist=|#mirrorlist=|g' \
-e 's|^#baseurl=http://dl.rockylinux.org/$contentdir|baseurl=https://mirrors.aliyun.com/rockylinux|g' \
-i.bak \
/etc/yum.repos.d/Rocky-*.repo
dnf makecache
关闭firewalld防火墙启用iptables
systemctl stop firewalld systemctl disable firewalld
yum -y install iptables-services systemctl start iptables iptables -F systemctl enable iptables service iptables save
禁用selinux
setenforce 0 sed -i "s/SELINUX=enforcing/SELINUX=disabled/g" /etc/selinux/config
关闭swap分区
swapoff -a sed -i 's:/dev/mapper/rl-swap:#/dev/mapper/rl-swap:g' /etc/fstab
修改主机名
hostnamectl set-hostname k8s-master # master hostnamectl set-hostname k8s-node01 # node01 hostnamectl set-hostname k8s-node02 # node02 bash # 刷新生效
配置本地解析
vi /etc/hosts 192.168.68.160 k8s-master m1 192.168.68.161 k8s-node01 n1 192.168.68.162 k8s-node02 n2
安装ipvs
yum install -y ipvsadm
开启路由转发
echo 'net.ipv4.ip_forward=1' >> /etc/sysctl.conf sysctl -p
加载 bridge
yum install -y epel-release yum install -y bridge-utils modprobe br_netfilter echo 'br_netfilter' >> /etc/modules-load.d/bridge.conf echo 'net.bridge.bridge-nf-call-iptables=1' >> /etc/sysctl.conf echo 'net.bridge.bridge-nf-call-ip6tables=1' >> /etc/sysctl.conf sysctl -p
安装docker
dnf config-manager --add-repo https://mirrors.ustc.edu.cn/docker-ce/linux/centos/docker-ce.repo # 切换中科大源 sed -e 's|download.docker.com|mirrors.ustc.edu.cn/docker-ce|g' /etc/yum.repos.d/docker-ce.repo # 安装docker yum -y install docker-ce # 或者使用阿里源 yum install -y yum-utils device-mapper-persistent-data lvm2 yum-config-manager --add-repo https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo sed -i 's+download.docker.com+mirrors.aliyun.com/docker-ce+' /etc/yum.repos.d/docker-ce.repo yum -y install docker-ce
配置docker
cat > /etc/docker/daemon.json <<EOF
{
"data-root": "/data/docker",
"exec-opts": ["native.cgroupdriver=systemd"],
"log-driver": "json-file",
"log-opts": {
"max-size": "100m",
"max-file": "100"
},
"insecure-registries": ["harbor.xinxainghf.com"],
"registry-mirrors": ["https://proxy.1panel.live","https://docker.1panel.top","https://docker.m.daocloud.io","https://docker.1ms.run","https://docker.ketches.cn"]
}
EOF
mkdir -p /etc/systemd/system/docker.service.d
systemctl daemon-reload && systemctl restart docker && systemctl enable docker
重启设备
reboot # 建议拍摄快照防止后续问题没办法恢复环境
安装cri-docker
yum -y install wget wget https://github.com/Mirantis/cri-dockerd/releases/download/v0.3.9/cri-dockerd-0.3.9.amd64.tgz
# 注意 下面操作必须 在配置本地解析时 n1 , n2 和我写的一样 cp cri-dockerd/cri-dockerd /usr/bin/ scp cri-dockerd/cri-dockerd root@n1:/usr/bin/cri-dockerd scp cri-dockerd/cri-dockerd root@n2:/usr/bin/cri-dockerd
配置 cri-docker 服务(这块仅在master设备上执行)
cat <<"EOF" > /usr/lib/systemd/system/cri-docker.service [Unit] Description=CRI Interface for Docker Application Container Engine Documentation=https://docs.mirantis.com After=network-online.target firewalld.service docker.service Wants=network-online.target Requires=cri-docker.socket [Service] Type=notify ExecStart=/usr/bin/cri-dockerd --network-plugin=cni --pod-infra-container-image=registry.aliyuncs.com/google_containers/pause:3.8 ExecReload=/bin/kill -s HUP $MAINPID TimeoutSec=0 RestartSec=2 Restart=always StartLimitBurst=3 StartLimitInterval=60s LimitNOFILE=infinity LimitNPROC=infinity LimitCORE=infinity TasksMax=infinity Delegate=yes KillMode=process [Install] WantedBy=multi-user.target EOF
cat <<"EOF" > /usr/lib/systemd/system/cri-docker.socket [Unit] Description=CRI Docker Socket for the API PartOf=cri-docker.service [Socket] ListenStream=%t/cri-dockerd.sock SocketMode=0660 SocketUser=root SocketGroup=docker [Install] WantedBy=sockets.target EOF
scp /usr/lib/systemd/system/cri-docker.service root@n1:/usr/lib/systemd/system/cri-docker.service scp /usr/lib/systemd/system/cri-docker.service root@n2:/usr/lib/systemd/system/cri-docker.service scp /usr/lib/systemd/system/cri-docker.socket root@n1:/usr/lib/systemd/system/cri-docker.socket scp /usr/lib/systemd/system/cri-docker.socket root@n2:/usr/lib/systemd/system/cri-docker.socket
启动cri-docker
systemctl daemon-reload systemctl enable cri-docker systemctl start cri-docker systemctl is-active cri-docker # 返回 active 即为成功
添加kuberadm源
https://v1-29.docs.kubernetes.io/zh-cn/docs/setup/production-environment/tools/kubeadm/install-kubeadm/ # 官方文档 参考使用
cat <<EOF > /etc/yum.repos.d/kubernetes.repo [kubernetes] name=Kubernetes baseurl=https://pkgs.k8s.io/core:/stable:/v1.29/rpm/ enabled=1 gpgcheck=1 gpgkey=https://pkgs.k8s.io/core:/stable:/v1.29/rpm/repodata/repomd.xml.key # exclude=kubelet kubeadm kubectl cri-tools kubernetes-cni EOF
安装kubeadm 1.29.2 版本
yum install -y kubelet-1.29.2 kubectl-1.29.2 kubeadm-1.29.2 systemctl enable kubelet.service
安装完成之后取消exclude的跳过注释,防止被update升级
sed -i 's/\# exclude/exclude/' /etc/yum.repos.d/kubernetes.repo
四、集群初始化
初始化master
# master 节点执行 kubeadm init --apiserver-advertise-address=192.168.68.160 --image-repository registry.aliyuncs.com/google_containers --kubernetes-version 1.29.2 --service-cidr=10.10.0.0/12 --pod-network-cidr=10.244.0.0/16 --ignore-preflight-errors=all --cri-socket unix:///var/run/cri-dockerd.sock
创建kubernetes证书目录
mkdir -p $HOME/.kube sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config sudo chown $(id -u):$(id -g) $HOME/.kube/config
node节点加入集群
# node 节点执行，复制集群初始化的token命令时间记得加 --cri-socket unix:///var/run/cri-dockerd.sock 指定cri-docker的sock文件 kubeadm join 192.168.68.160:6443 --token kvuy4b.s4hanac9sc2g9uf5 --discovery-token-ca-cert-hash sha256:0cd92bb4735b5e8444bd58ae9b72eef16888e03e0768323f280cbaf0d711c094 --cri-socket unix:///var/run/cri-dockerd.sock
验证node节点是否加入集群
kubernetes get node
如果 token过期 重新申请
kubeadm token create --print-join-command
五、部署网络calico插件
官方文档
https://docs.tigera.io/calico/latest/getting-started/kubernetes/self-managed-onprem/onpremises#install-calico-with-kubernetes-api-datastore-more-than-50-nodes
导入镜像
# master 节点执行 wget https://alist.wanwancloud.cn/d/%E8%BD%AF%E4%BB%B6/Kubernetes/Calico/calico.zip?sign=45yzzkju25P3oo3e7-D7w4A38_0Ug50Xag585imOt-0=:0 -O calico.zip yum -y install unzip unzip calico.zip cd calico tar xf calico-images.tar.gz scp -r calico-images root@n1:/root scp -r calico-images root@n2:/root docker load -i calico-images/calico-cni-v3.26.3.tar docker load -i calico-images/calico-kube-controllers-v3.26.3.tar docker load -i calico-images/calico-node-v3.26.3.tar docker load -i calico-images/calico-typha-v3.26.3.tar
导入镜像
# node 节点执行 docker load -i calico-images/calico-cni-v3.26.3.tar docker load -i calico-images/calico-kube-controllers-v3.26.3.tar docker load -i calico-images/calico-node-v3.26.3.tar docker load -i calico-images/calico-typha-v3.26.3.tar
修改yaml文件
vi calico-typha.yaml
CALICO_IPV4POOL_CIDR 指定为 pod 地址 也是就是初始化集群时的地址 --pod-network-cidr=10.244.0.0/16 - name: CALICO_IPV4POOL_CIDR value: "10.244.0.0/16"
CALICO_IPV4POOL_IPIP 修改为 BGP 模式 - name: CALICO_IPV4POOL_IPIP value: "Always" #改成Off
启用calico网络插件
kubectl apply -f calico-typha.yaml
稍等一段时间 组件全部 Running 集群节点 Ready 就绪 集群就搭建成功了
kubectl get pod -A kubectl get node
资源清单
一、资源清单详解
[root@k8s-master ~]# kubectl api-versions # 获取版本 admissionregistration.k8s.io/v1 apiextensions.k8s.io/v1 apiregistration.k8s.io/v1 apps/v1 authentication.k8s.io/v1 authorization.k8s.io/v1 autoscaling/v1 autoscaling/v2 batch/v1 certificates.k8s.io/v1 coordination.k8s.io/v1 crd.projectcalico.org/v1 discovery.k8s.io/v1 events.k8s.io/v1 flowcontrol.apiserver.k8s.io/v1 flowcontrol.apiserver.k8s.io/v1beta3 networking.k8s.io/v1 node.k8s.io/v1 policy/v1 rbac.authorization.k8s.io/v1 scheduling.k8s.io/v1 storage.k8s.io/v1 v1
# Pod 资源清单 详解
apiVersion: v1 # 接口组版本 可通过 kubectl explain pod 获取 core 核心v1 版本 简写v1
kind: Pod # 资源类型 Pod
metadata: # 元数据
name: pod-demo # pod 名字
namespace: default # 所在命名空间 不写也默认 default
labels: # 标签选择
app: myapp # 定义标签 便于归类识别等
spec: # 期望 也就是你最终想让他帮你达成的状态 最终能不能成功是不一定的 也属于声明式的表达
# 例如我想让他帮我创建一个不存在镜像的pod，肯定创建不出来但是他会一直帮你创建达到你的期望状态
# 故此这里 spec 不叫定义而是期望
containers: # 期望子对象 容器组
- name: myapp-1 # 第一个容器名 myapp-1
image: wangyanglinux/myapp:v1.0 # 容器镜像
- name: buxybox-1 # 第二个容器名 buxybox-1
image: wangyanglinux/tools:busybox # 容器镜像
command: # buxybox-1 容器的对应运行命令
- "/bin/sh"
- "-c"
- "sleep 3600"
status: # 状态 描绘这个资源清单的状态
conditions:
- lastProbeTime: null
lastTransitionTime: "2024-12-20T08:49:07Z"
status: "True"
type: Initialized
- lastProbeTime: null
二、创建资源
编写资源清单
[root@k8s-master ~]# cat >> pod.yml <<EOF
apiVersion: v1
kind: Pod
metadata:
name: pod-demo
namespace: default
labels:
app: myapp
spec:
containers:
- name: myapp-1
image: wangyanglinux/myapp:v1.0
- name: buxybox-1
image: wangyanglinux/tools:busybox
command:
- "/bin/sh"
- "-c"
- "sleep 3600"
status:
conditions:
- lastProbeTime: null
lastTransitionTime: "2024-12-20T08:49:07Z"
status: "True"
type: Initialized
- lastProbeTime: null
EOF
把期望资源清单提交k8s集群，帮我们进行对应的生产制造 进行实例化
[root@k8s-master ~]# kubectl create -f pod.yml
[root@k8s-master ~]# kubectl get pod # 查看 pod 状态 NAME READY STATUS RESTARTS AGE pod-demo 0/2 ContainerCreating 0 5s [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE pod-demo 2/2 Running 0 19s [root@k8s-master ~]# kubectl get pod -o wide # 查看 pod 详细信息 NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES pod-demo 2/2 Running 0 28s 10.244.85.201 k8s-node01 <none> <none>
根据上面详细信息 NODE k8s-node01 可以获取到当前 pod 的所在节点
登陆 nod01 设备执行命令 获取容器信息 可以得出概念 pod 是由多个容器组成在一起的 pause k8s_myapp k8s_buxybox 三个容器构建成的 pod [root@k8s-node01 ~]# docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES f9a71e2a9bb1 adf0836b2bab "/bin/sh -c 'sleep 3…" About a minute ago Up About a minute k8s_buxybox-1_pod-demo_default_37ceeb28-143a-4f02-9f2f-6cfdbe25dc1f_0 df23f643af98 79fbe47c0ab9 "/bin/sh -c 'hostnam…" About a minute ago Up About a minute k8s_myapp-1_pod-demo_default_37ceeb28-143a-4f02-9f2f-6cfdbe25dc1f_0 21b1d02515a3 registry.aliyuncs.com/google_containers/pause:3.8 "/pause" About a minute ago Up About a minute k8s_POD_pod-demo_default_37ceeb28-143a-4f02-9f2f-6cfdbe25dc1f_0
三、修改Pod内容器信息
[root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES pod-demo 2/2 Running 0 28s 10.244.85.201 k8s-node01 <none> <none> [root@k8s-master ~]# curl 10.244.85.201 # 访问 pod IP 得到下面信息 www.xinxianghf.com | hello MyAPP | version v1.0
** 使用kubectl 修改容器内部信息**
# 命令解析 exec 进入容器 -it 分配可交互的终端 pod-demo pod的名字 -c 指定 myapp 容器名 -- 分割 /bin/bash 执行的命令 [root@k8s-master ~]# kubectl exec -it pod-demo -c myapp-1 -- /bin/bash pod-demo:/# cd /usr/local/nginx/html/ pod-demo:/usr/local/nginx/html# cd /usr/local/nginx/html/ pod-demo:/usr/local/nginx/html# echo "123321" >> index.html pod-demo:/usr/local/nginx/html# exit exit
追加的内容已经生效
四、标签查看
[root@k8s-master ~]# kubectl get pod --show-labels # 查看default命名空间下的所有标签 NAME READY STATUS RESTARTS AGE LABELS pod-demo 2/2 Running 0 14m app=myapp [root@k8s-master ~]# kubectl get pod -l app # 查看default命名空间下具备app这个key的标签的pod NAME READY STATUS RESTARTS AGE pod-demo 2/2 Running 0 14m [root@k8s-master ~]# kubectl get pod -l app=myapp # 查看default命名空间下具备app=myapp标签的pod NAME READY STATUS RESTARTS AGE pod-demo 2/2 Running 0 15m [root@k8s-master ~]# kubectl get pod -l app1 # 查看具备app1这个key的标签的pod 没有app1所以不显示 No resources found in default namespace
五、查看容器日志
# pod-demo pod名 -c 指定 容器名 [root@k8s-master ~]# kubectl logs pod-demo -c myapp-1 192.168.68.160 - - [20/Dec/2024:14:32:51 +0800] "GET / HTTP/1.1" 200 48 "-" "curl/7.76.1" 192.168.68.160 - - [20/Dec/2024:14:42:31 +0800] "GET / HTTP/1.1" 200 55 "-" "curl/7.76.1" # 如果 pod 内只有一个容器 那么不需要-c 指定容器名
六、查看资源对象详细信息
[root@k8s-master ~]# cat >> 2.pod.yml <<EOF apiVersion: v1 kind: Pod metadata: name: pod-demo-1 namespace: default labels: app: myapp spec: containers: - name: myapp-1 image: wangyanglinux/myapp:v1.0 - name: myapp-2 image: wangyanglinux/myapp:v2.0 EOF
kubectl create -f 2.pod.yml
kubectl discribe pod pod-demo-1
七、Pod的特性
续接6板块 发现pod特性
kubectl discribe pod pod-demo-1
# 这里会发现 容器只就绪了一个并且创建6次都失败的 [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE pod-demo-1 1/2 CrashLoopBackOff 6 (17s ago) 6m41s
查看具体的容器日志
[root@k8s-master ~]# kubectl logs pod-demo-1 -c myapp-1 [root@k8s-master ~]# kubectl logs pod-demo-1 -c myapp-2 # 发现容器2 的80端口显示已经被占用 2024/12/20 15:17:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use) nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use) 2024/12/20 15:17:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use) nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use) 2024/12/20 15:17:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use) nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use) 2024/12/20 15:17:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use) nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use) 2024/12/20 15:17:40 [emerg] 1#1: bind() to 0.0.0.0:80 failed (98: Address in use) nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address in use) 2024/12/20 15:17:40 [emerg] 1#1: still could not bind() nginx: [emerg] still could not bind()
# 镜像wangyanglinux/myapp:v1.0 是基于操作系统封装编译安装的nginx # 由此可以得出 同一个pod内部 不同的容器中共享网络站 IPC PID # 当容器1启动后占用80端口 容器2就无法继续使用被占用的80端口
八、常用命令
# 获取当前的资源 pod
$ kubectl get pod
-A , --all-namespaces 查看所有名称空间下的资源
-n 指定空间名称，不写默认default
--show-labels 查看当前标签
-l 筛选资源 key、key=value
-o wide 详细信息 分配的节点设备 ip等
# 进入pod执行命令 pod名 容器名
$ kubectl exec -it podName -c cName -- command
-c 可以省略 默认进入容器为一的容器内部
# 查看资源的描述
$ kubectl explain pod.spec
# 查看容器内的日志
$ kubectl logs podName -c cName
-c 可以省略 默认查看容器为一的容器日志
# 创建 pod
$ kubectl create -f pod.yml
# 删除 pod
$ kubectl delete pod pod-demo # 通过 pod 名称删除对应 pod
$ kubectl delete -f pod.yml # 通过yml文件删除对应资源
# 查看资源对象的详细描述
$ kubectl describe pod podName
Pod的生命周期
从创建到死亡的完成过程
Pod控制器-核心灵魂
在Kubernetes中运行了一系列控制器来确保集群的当前状态与期望状态保持一致，它们就是Kubernetes集群内部的管理控制中心或者说是“中心大脑” 例如： ReplicaSet控制器负责维护集群中运行的Pod数量 节点控制器负责监控节点的状态，并在节点出现故障时，执行自动化修复流程，确保集群始终处于预期的工作状态。
一、ReplicationController控制器
简写RC 用来确保容器应用的副本数量始终保持在用户定义的副本书，即如果有容器异常退出，会自动创建新的 Pod 来替代；而如果异常多出来的容器也会自动回收； 在新版的 Kubernetes 中建议使用 ReplicaSet 来取代 ReplicationController。 ReplicaSet 跟 ReplicationController 没有本质的不同，知识名字不一样，而且 ReplicaSet 支持集合式的selector。
# 解释 apiVersion: v1 # 核心组 v1版本 kind: ReplicationController # 资源类别 ReplicationController 控制器 metadata: # 元数据 name: rc-demo # rc 控制器名 spec: # 期望 replicas: 3 # pod 副本数量 selector: # 选择器 app: rc-demo # rc 控制器标签 template: # 模版 metadata: # 元数据 labels: # Pod 标签 app: rc-demo # 必须是rc 控制器标签的子集 spec: # 容器组的期望 containers: # 容器组 - name: rc-demo-container # 容器名 image: wangyanglinux/myapp:v1.0 # 容器镜像 env: # 环境变量 - name: GET_HOSTS_FROM # 定义环境变量 value: dns name: zhangsan value: "123" ports: # 容器端口 - containerPort: 80 # 80端口 EOF
[root@k8s-master ~]# cat >> rc.yml <<EOF apiVersion: v1 kind: ReplicationController metadata: name: rc-demo spec: replicas: 3 selector: app: rc-demo template: metadata: labels: app: rc-demo spec: containers: - name: rc-demo-container image: wangyanglinux/myapp:v1.0 env: - name: GET_HOSTS_FROM value: dns name: zhangsan value: "123" ports: - containerPort: 80 EOF
[root@k8s-master ~]# kubectl create -f rc.yml
通过删除 pod 的方式验证
[root@k8s-master ~]# kubectl get rc # 查看rc控制器资源 NAME DESIRED CURRENT READY AGE rc-demo 3 3 3 6m26s # 副本3 创建3 就绪3 [root@k8s-master ~]# kubectl get pod # 查看 pod 资源 NAME READY STATUS RESTARTS AGE rc-demo-ndx2r 1/1 Running 0 6m32s rc-demo-nhv9q 1/1 Running 0 6m32s rc-demo-rtlph 1/1 Running 0 6m32s [root@k8s-master ~]# kubectl delete pod rc-demo-ndx2r # 手动删除 pod pod "rc-demo-ndx2r" deleted [root@k8s-master ~]# kubectl get pod # 查看 pod 资源 NAME READY STATUS RESTARTS AGE rc-demo-ljdhd 1/1 Running 0 11s # 发现原删除的 pod 自动重新创建了 rc-demo-nhv9q 1/1 Running 0 7m28s rc-demo-rtlph 1/1 Running 0 7m28s
通过标签的子集方式验证
[root@k8s-master ~]# kubectl get pod --show-labels # 通过标签查看pod信息 NAME READY STATUS RESTARTS AGE LABELS rc-demo-ljdhd 1/1 Running 0 9m58s app=rc-demo rc-demo-nhv9q 1/1 Running 0 17m app=rc-demo rc-demo-rtlph 1/1 Running 0 17m app=rc-demo [root@k8s-master ~]# kubectl label pod rc-demo-ljdhd version=v1 # 新增rc-demo-ljdhd 这个pod 标签 pod/rc-demo-ljdhd labeled [root@k8s-master ~]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS rc-demo-ljdhd 1/1 Running 0 10m app=rc-demo,version=v1 # 增加了version=v1的标签，但这个pod的还是rc控制器标签的子集 rc-demo-nhv9q 1/1 Running 0 17m app=rc-demo rc-demo-rtlph 1/1 Running 0 17m app=rc-demo
[root@k8s-master ~]# kubectl get pod --show-labels # 通过标签查看pod信息 NAME READY STATUS RESTARTS AGE LABELS rc-demo-ljdhd 1/1 Running 0 10m app=rc-demo,version=v1 rc-demo-nhv9q 1/1 Running 0 17m app=rc-demo rc-demo-rtlph 1/1 Running 0 17m app=rc-demo [root@k8s-master ~]# kubectl label pod rc-demo-nhv9q app=test --overwrite # 修改app标签为test pod/rc-demo-nhv9q labeled [root@k8s-master ~]# kubectl get pod --show-labels # 通过标签查看pod信息 发现现在是4个pod NAME READY STATUS RESTARTS AGE LABELS rc-demo-jdnfm 1/1 Running 0 66s app=rc-demo # 新建的pod rc-demo-ljdhd 1/1 Running 0 15m app=rc-demo,version=v1 rc-demo-nhv9q 1/1 Running 0 23m app=test # app=test 不符合 rc控制器标签的子集 故不在受rc控制器管理 rc-demo-rtlph 1/1 Running 0 23m app=rc-demo
[root@k8s-master ~]# kubectl get pod --show-labels # 通过标签查看pod信息 NAME READY STATUS RESTARTS AGE LABELS rc-demo-jdnfm 1/1 Running 0 66s app=rc-demo rc-demo-ljdhd 1/1 Running 0 15m app=rc-demo,version=v1 rc-demo-nhv9q 1/1 Running 0 23m app=test rc-demo-rtlph 1/1 Running 0 23m app=rc-demo [root@k8s-master ~]# kubectl label pod rc-demo-nhv9q app=rc-demo --overwrite # 修改app标签为rc-demo pod/rc-demo-nhv9q labeled [root@k8s-master ~]# kubectl get pod --show-labels # 通过过标签查看pod信息 发现现在是3个pod # 最新创建的rc-demo-jdnfm这个pod已经被删除了，由此可见符合 rc控制器标签都将被rc控制器所管理 NAME READY STATUS RESTARTS AGE LABELS rc-demo-ljdhd 1/1 Running 0 18m app=rc-demo,version=v1 rc-demo-nhv9q 1/1 Running 0 25m app=rc-demo rc-demo-rtlph 1/1 Running 0 25m app=rc-demo
由此可见 rc控制器最重要的特性： - 保证副本值与期望值尽可能的一致 - 当然也不可能会完全一致，如果node资源不够无法创建指定数量的pod你那么他会一直的重试，尽可能的达到你期望值的数量 - pod副本数量多出会优先杀死删除最新创建的pod - pod副本数量少会创建新pod - rc控制器通过标签管理pod的数量
通过命令调整rc副本的数量
[root@k8s-master ~]# kubectl get pod -o wide # 获取当前的pod NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES rc-demo-ljdhd 1/1 Running 0 24m 10.244.85.206 k8s-node01 <none> <none> rc-demo-nhv9q 1/1 Running 0 31m 10.244.85.205 k8s-node01 <none> <none> rc-demo-rtlph 1/1 Running 0 31m 10.244.58.200 k8s-node02 <none> <none> [root@k8s-master ~]# kubectl scale rc rc-demo --replicas=10 # 调整名称为 rc-demo 的 rc 控制器 pod 副本数量为10 replicationcontroller/rc-demo scaled [root@k8s-master ~]# kubectl get pod -o wide # 获取 pod 信息 NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES rc-demo-4l22s 1/1 Running 0 39s 10.244.85.211 k8s-node01 <none> <none> rc-demo-52mg4 1/1 Running 0 39s 10.244.58.205 k8s-node02 <none> <none> rc-demo-57z42 1/1 Running 0 39s 10.244.58.208 k8s-node02 <none> <none> rc-demo-8mzzt 1/1 Running 0 39s 10.244.58.207 k8s-node02 <none> <none> rc-demo-gh96h 1/1 Running 0 39s 10.244.58.206 k8s-node02 <none> <none> rc-demo-khxrw 1/1 Running 0 39s 10.244.85.213 k8s-node01 <none> <none> rc-demo-ljdhd 1/1 Running 0 25m 10.244.85.206 k8s-node01 <none> <none> rc-demo-nhv9q 1/1 Running 0 32m 10.244.85.205 k8s-node01 <none> <none> rc-demo-rtlph 1/1 Running 0 32m 10.244.58.200 k8s-node02 <none> <none> rc-demo-zdzlt 1/1 Running 0 39s 10.244.85.212 k8s-node01 <none> <none> [root@k8s-master ~]# kubectl get rc # 获取 rc 控制器的信息 NAME DESIRED CURRENT READY AGE rc-demo 10 10 10 33m # 副本10 创建10 就绪10
[root@k8s-master ~]# kubectl scale rc rc-demo --replicas=2 # 调整名称为 rc-demo 的 rc 控制器 pod 副本数量为10 replicationcontroller/rc-demo scaled [root@k8s-master ~]# kubectl get rc # 获取 rc 控制器信息 NAME DESIRED CURRENT READY AGE rc-demo 2 2 2 35m [root@k8s-master ~]# kubectl get pod -o wide # 获取 pod 数量 NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES rc-demo-nhv9q 1/1 Running 0 35m 10.244.85.205 k8s-node01 <none> <none> rc-demo-rtlph 1/1 Running 0 35m 10.244.58.200 k8s-node02 <none> <none>
二、ReplicaSet 控制器
# 解释 apiVersion: apps/v1 # rs 控制器核心组版本 kind: ReplicaSet # 资源类型 rs 控制器 metadata: # 元数据 name: rs-ml-demo # rs 控制器名 spec: # 期望 replicas: 3 # 副本数量 selector: # 标签选择 matchLabels: # 使用子集匹配标签跟 pod 标签做子集运算 app: rs-ml-demo # 定义rs 控制器标签 domain: rs-t1 # 定义rs 控制器标签 template: # 定义pod模版 metadata: # 元数据 labels: # pod 标签 app: rs-ml-demo # 定义 pod 标签 domain: rs-t1 version: v1 spec: # 容器期望 containers: # 容器组 - name: rs-ml-demo-container # 容器名 image: wangyanglinux/myapp:v1.0 # 容器镜像 env: # 定义容器内环境变量 - name: GET_HOSTS_FROM value: dns ports: # 声明端口 可写可不写 - containerPort: 80 # 容器使用端口 80
[root@k8s-master ~]# cat >> rs.yml <<EOF
apiVersion: apps/v1
kind: ReplicaSet
metadata:
name: rs-ml-demo
spec:
replicas: 3
selector:
matchLabels:
app: rs-ml-demo
domain: rs-t1
template:
metadata:
labels:
app: rs-ml-demo
domain: rs-t1
version: v1
spec:
containers:
- name: rs-ml-demo-container
image: wangyanglinux/myapp:v1.0
env:
- name: GET_HOSTS_FROM
value: dns
ports:
- containerPort: 80
EOF
[root@k8s-master ~]# kubectl create -f rs.yml
[root@k8s-master ~]# kubectl get pod # 查看 pod 信息 NAME READY STATUS RESTARTS AGE rs-ml-demo-4shl5 1/1 Running 0 2m12s rs-ml-demo-p9hbt 1/1 Running 0 2m12s rs-ml-demo-qs9dw 1/1 Running 0 2m12s [root@k8s-master ~]# kubectl get rs # 查看 rs 控制器信息 NAME DESIRED CURRENT READY AGE rs-ml-demo 3 3 3 2m57s [root@k8s-master ~]# kubectl delete pod rs-ml-demo-4shl5 # 删除 pod pod "rs-ml-demo-4shl5" deleted [root@k8s-master ~]# kubectl get pod # 查看 pod 信息 NAME READY STATUS RESTARTS AGE rs-ml-demo-7bhfg 1/1 Running 0 12s # 删除的 pod 被 rs 自动重新创建 rs-ml-demo-p9hbt 1/1 Running 0 3m41s rs-ml-demo-qs9dw 1/1 Running 0 3m41s
matchExpressions 标签
selector.matchExpressions rs 在标签选择器上，除了可以定义键值对的选择形式，还支持 matchExpressions 字段，可以提供多种选择。 目前支持的操作包括： - In：label 的值在某个列表中 - NotIn：label 的值不在某个列表中 - Exists：某个 label 存在 - DoesNotExist：某个 label 不存在
# 解释 apiVersion: apps/v1 kind: ReplicaSet metadata: name: rs-me-exists-demo spec: selector: matchExpressions: # 匹配运算标签 - key: app # 标签app operator: Exists # 包含 app 这个key template: metadata: labels: app: spring-k8s # pod 标签 spec: containers: - name: rs-me-exists-demo-container image: wangyanglinux/myapp:v1.0 ports: - containerPort: 80
[root@k8s-master ~]# cat >> rs-1.yml <<EOF apiVersion: apps/v1 kind: ReplicaSet metadata: name: rs-me-exists-demo spec: selector: matchExpressions: - key: app operator: Exists template: metadata: labels: app: spring-k8s spec: containers: - name: rs-me-exists-demo-container image: wangyanglinux/myapp:v1.0 ports: - containerPort: 80 EOF
kubectl create -f rs-1.yml
[root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE rs-me-exists-demo-v245q 1/1 Running 0 3m12s rs-ml-demo-7bhfg 1/1 Running 0 23m rs-ml-demo-p9hbt 1/1 Running 0 26m rs-ml-demo-qs9dw 1/1 Running 0 26m [root@k8s-master ~]# kubectl get rs # 资源控制器未写副本数量 默认为1 NAME DESIRED CURRENT READY AGE rs-me-exists-demo 1 1 1 5m45s rs-ml-demo 3 3 3 29m [root@k8s-master ~]# kubectl get pod --show-labels # 查看当前 pod 标签 NAME READY STATUS RESTARTS AGE LABELS rs-me-exists-demo-v245q 1/1 Running 0 3m20s app=spring-k8s # rs exists key匹配创建的 pod rs-ml-demo-7bhfg 1/1 Running 0 23m app=rs-ml-demo,domain=rs-t1,version=v1 rs-ml-demo-p9hbt 1/1 Running 0 26m app=rs-ml-demo,domain=rs-t1,version=v1 rs-ml-demo-qs9dw 1/1 Running 0 26m app=rs-ml-demo,domain=rs-t1,version=v1 [root@k8s-master ~]# kubectl get rs NAME DESIRED CURRENT READY AGE rs-me-exists-demo 1 1 1 3m46s rs-ml-demo 3 3 3 27m [root@k8s-master ~]# kubectl label pod rs-me-exists-demo-v245q app=test-1 --overwrite # 修改标签的 value 为test-1 pod/rs-me-exists-demo-v245q labeled [root@k8s-master ~]# kubectl get pod --show-labels # 查看当前 pod 标签 NAME READY STATUS RESTARTS AGE LABELS rs-me-exists-demo-v245q 1/1 Running 0 5m40s app=test-1 # 发现pod并未被删除 rs-ml-demo-7bhfg 1/1 Running 0 25m app=rs-ml-demo,domain=rs-t1,version=v1 rs-ml-demo-p9hbt 1/1 Running 0 29m app=rs-ml-demo,domain=rs-t1,version=v1 rs-ml-demo-qs9dw 1/1 Running 0 29m app=rs-ml-demo,domain=rs-t1,version=v1
# 得出结论 使用 matchExpressions Exists 标签时 key做匹配 value不做匹配
# rs控制器就可管理携带key标签一致的pod，但是不会管理到其他非该资源创建出来的pod
# rs 控制器比 Rc控制器 有更丰富的 标签选择管理功能
spec:
selector:
matchExpressions:
- key: app # 标签key = app
operator: Exists
template:
metadata:
labels:
app: spring-k8s # pod 标签的app = app 所以可以被rs控制器管理
三、Deployment控制器
基本概念
声明式与命令式
声明性的东西是对终结果的陈述，表明意图而不是实现它的过程。在Kubernetes中，这就是说"应该有一个包含三个Pod的ReplicaSet" 命令式充当命令。声明式是被动的，而命令是主动且直接的："创建一个包含三个Pod的ReplicaSet"。
replace 与apply 的对比 替换方式 kubectl replace:使用新的配置完全替换掉现有资源的配置。这意味着新配置将覆盖现有资源的所有字段和属性，包括未指定的字段，会导致整个资源的替换 kubectl apply: 使用新的配置部分地更新现有资源的配置。它会根据提供的配置文件或参数，只更新与新配置中不同的部分，而不会覆盖整个资源的配置字段级别的更新 kubectl replace: 由于是完全替换，所以会覆盖所有字段和属性，无论是否在新配置中指定 kubectl apply:只更新与新配置中不同的字段和属性，保留未指定的字段不受影响 部分更新 kubectl replace: 不支持部分更新，它会替换整个资源的配置 kubectl apply: 支持部分更新，只会更新新配置中发生变化的部分，保留未指定的部分不受影响 与其他配置的影响 kubectl replace: 不考虑其他资源配置的状态，直接替换资源的配置 kubectl apply: 可以结合使用 -f或 -k 参数，从文件或目录中读取多个资源配置，并根据当前集群中的资源状态进行更新
# --dry-run -o yaml 不创建资源生成模版
[root@k8s-master ~]# kubectl create deployment myapp --image=wangyanglinux/myapp:v1.0 --dry-run -o yaml > deployment.yml.tmp
W1224 11:38:42.790200 874724 helpers.go:704] --dry-run is deprecated and can be replaced with --dry-run=client.
[root@k8s-master ~]# cat deployment.yml.tmp
apiVersion: apps/v1
kind: Deployment
metadata:
creationTimestamp: null
labels:
app: myapp
name: myapp
spec:
replicas: 1
selector:
matchLabels:
app: myapp
strategy: {}
template:
metadata:
creationTimestamp: null
labels:
app: myapp
spec:
containers:
- image: wangyanglinux/myapp:v1.0
name: myapp
resources: {}
status: {}
# 解释 apiVersion: apps/v1 # Deployment资源核心版本 kind: Deployment # 资源类别 metadata: # 元数据 labels: # deployment标签 app: deployment-demo # 定义标签 name: deployment-demo # 控制器名字 spec: # 期望 selector: # deployment 标签 matchLabels: # 使用的匹配标签 app: deployment-demo # 定义标签 key 和 value template: # 副本 metadata: # 元数据 labels: # Pod 标签 app: deployment-demo # 定义pod标签 key value spec: # 期望 containers: # 容器组 - image: wangyanglinux/myapp:v1.0 # 镜像 name: deployment-demo-container # 容器名
[root@k8s-master ~]# cat >> deployment.yml << EOF apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v1.0 name: deployment-demo-container EOF
[root@k8s-master ~]# kubectl apply -f deployment.yml # 创建资源 deployment.apps/deployment-demo created [root@k8s-master ~]# kubectl get deploy # 查看 deploy 资源控制器 NAME READY UP-TO-DATE AVAILABLE AGE deployment-demo 1/1 1 1 30s [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-6995c75668-5g8p7 1/1 Running 0 70s
查看 deployment-demo 控制器存储在etcd数据库的数据
[root@k8s-master ~]# kubectl get deployment deployment-demo -o yaml
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment","metadata":{"annotations":{},"labels":{"app":"deployment-demo"},"name":"deployment-demo","namespace":"default"},"spec":{"replicas":1,"selector":{"matchLabels":{"app":"deployment-demo"}},"template":{"metadata":{"labels":{"app":"deployment-demo"}},"spec":{"containers":[{"image":"wangyanglinux/myapp:v2.0","name":"deployment-demo-container"}]}}}}
creationTimestamp: "2024-12-20T12:48:41Z"
generation: 1
labels:
app: deployment-demo
name: deployment-demo
namespace: default
resourceVersion: "394942"
uid: da19be8e-1d44-4855-b981-84b142f681d7
spec:
progressDeadlineSeconds: 600
replicas: 1
revisionHistoryLimit: 10
selector:
matchLabels:
app: deployment-demo
strategy:
rollingUpdate:
maxSurge: 25%
maxUnavailable: 25%
type: RollingUpdate
template:
metadata:
creationTimestamp: null
labels:
app: deployment-demo
spec:
containers:
- image: wangyanglinux/myapp:v1.0
imagePullPolicy: IfNotPresent
name: deployment-demo-container
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
dnsPolicy: ClusterFirst
restartPolicy: Always
schedulerName: default-scheduler
securityContext: {}
terminationGracePeriodSeconds: 30
status:
availableReplicas: 1
conditions:
- lastTransitionTime: "2024-12-20T12:48:58Z"
lastUpdateTime: "2024-12-20T12:48:58Z"
message: Deployment has minimum availability.
reason: MinimumReplicasAvailable
status: "True"
type: Available
- lastTransitionTime: "2024-12-20T12:48:41Z"
lastUpdateTime: "2024-12-20T12:48:58Z"
message: ReplicaSet "deployment-demo-6465d4c5c9" has successfully progressed.
reason: NewReplicaSetAvailable
status: "True"
type: Progressing
observedGeneration: 1
readyReplicas: 1
replicas: 1
updatedReplicas: 1
# 增加副本数量 [root@k8s-master ~]# kubectl scale deployment deployment-demo --replicas=10 # 设置deployment-demo控制器pod副本数量10 deployment.apps/deployment-demo scaled [root@k8s-master ~]# kubectl get pod -o wide # 查看pod详细信息 NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6995c75668-7jvfq 1/1 Running 0 66s 10.244.85.238 k8s-node01 <none> <none> deployment-demo-6995c75668-94fhp 1/1 Running 0 65s 10.244.85.239 k8s-node01 <none> <none> deployment-demo-6995c75668-9v2dx 1/1 Running 0 66s 10.244.58.231 k8s-node02 <none> <none> deployment-demo-6995c75668-cscv9 1/1 Running 0 2m3s 10.244.85.236 k8s-node01 <none> <none> deployment-demo-6995c75668-gr2l6 1/1 Running 0 65s 10.244.58.232 k8s-node02 <none> <none> deployment-demo-6995c75668-l67tk 1/1 Running 0 66s 10.244.58.229 k8s-node02 <none> <none> deployment-demo-6995c75668-m4hqc 1/1 Running 0 66s 10.244.85.240 k8s-node01 <none> <none> deployment-demo-6995c75668-qq587 1/1 Running 0 66s 10.244.58.228 k8s-node02 <none> <none> deployment-demo-6995c75668-s6wxf 1/1 Running 0 66s 10.244.58.230 k8s-node02 <none> <none> deployment-demo-6995c75668-vdfgg 1/1 Running 0 66s 10.244.85.237 k8s-node01 <none> <none> [root@k8s-master ~]# curl 10.244.85.239 # 访问pod ip www.xinxianghf.com | hello MyAPP | version v1.0
版本升级 把镜像版本改为v2.0
[root@k8s-master ~]# sed -i 's/myapp:v1.0/myapp:v2.0/g' deployment.yml
[root@k8s-master ~]# kubectl apply -f deployment.yml # apply 更新 [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-4j898 1/1 Running 0 2m22s 10.244.58.234 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-4wccv 1/1 Running 0 87s 10.244.85.245 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-592z4 1/1 Running 0 99s 10.244.58.236 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-bzwxz 1/1 Running 0 2m25s 10.244.85.241 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-gbnb8 1/1 Running 0 2m24s 10.244.58.233 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-j5vh9 1/1 Running 0 2m21s 10.244.85.243 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-jr4m8 1/1 Running 0 2m24s 10.244.85.242 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-k7dm6 1/1 Running 0 91s 10.244.58.237 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-zqb42 1/1 Running 0 104s 10.244.58.235 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-zrhbs 1/1 Running 0 92s 10.244.85.244 k8s-node01 <none> <none> [root@k8s-master ~]# curl 10.244.58.234 # 更新为v2.0版本 www.xinxianghf.com | hello MyAPP | version v2.0
更新的过程中也发现一个新的特性，更新流程 先创建新版本的pod runing 后才删除旧版本的pod apply：只改变你声明的跟现在运行状态不相符的部分 其他不做改变 例如通过scale重新定义deployment-demo副本的数量为10，通过apply 重新更新的pod数量还是为10个
通过 replace 构建资源
[root@k8s-master ~]# sed -i 's/myapp:v2.0/myapp:v3.0/g' deployment.yml # 重新设置版本为v3.0 [root@k8s-master ~]# kubectl replace -f deployment.yml # 通过 replace 构建资源 deployment.apps/deployment-demo replaced [root@k8s-master ~]# kubectl get pod -o wide # 查看pod信息 数量只为1了 NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-7dbccb74d6-twd84 1/1 Running 0 111s 10.244.85.246 k8s-node01 <none> <none> [root@k8s-master ~]# kubectl get deployment # 查看控制器信息副本数量也只为1了 NAME READY UP-TO-DATE AVAILABLE AGE deployment-demo 1/1 1 1 40m [root@k8s-master ~]# curl 10.244.85.246 # 版本也已经更新 www.xinxianghf.com | hello MyAPP | version v3.0
得出结论 replace：会根据 deployment.yml 完全重建，完全覆盖 并不是更新与先版本不一致的地方 上面 副本数量为1 是因为deployment.yml 没有写副本数量，即默认为1 故此通过replace创建的pod的数量也为1
通过diff 对比当前现有deployment.yml的资源清单与已经创建的运行的资源对象做对比查询是否有区别
[root@k8s-master ~]# kubectl diff -f deployment.yml # 这里无返回结果 说明创建的资源与yml一致
[root@k8s-master ~]# sed -i 's/myapp:v3.0/myapp:v4.0/g' deployment.yml # 修改容器镜像版本号
[root@k8s-master ~]# kubectl diff -f deployment.yml # 再次对比
diff -u -N /tmp/LIVE-4200347792/apps.v1.Deployment.default.deployment-demo /tmp/MERGED-2776630918/apps.v1.Deployment.default.deployment-demo
--- /tmp/LIVE-4200347792/apps.v1.Deployment.default.deployment-demo 2024-12-20 21:34:15.119335901 +0800
+++ /tmp/MERGED-2776630918/apps.v1.Deployment.default.deployment-demo 2024-12-20 21:34:15.121335985 +0800
@@ -4,7 +4,7 @@
annotations:
deployment.kubernetes.io/revision: "6"
creationTimestamp: "2024-12-20T12:48:41Z"
- generation: 9
+ generation: 10
labels:
app: deployment-demo
name: deployment-demo
@@ -30,7 +30,7 @@
app: deployment-demo
spec:
containers:
- - image: wangyanglinux/myapp:v3.0 # 区别如下 删除了3.0版本的这一行
+ - image: wangyanglinux/myapp:v4.0 # 新增了4.0版本的这行
imagePullPolicy: IfNotPresent
name: deployment-demo-container
resources: {}
tips： 正式的生产环境中 可以使用diff 命令检测这个资源清单文件是否已经应用过，无反馈结果就是已经应用，有返回不同的地方说明这个资源清单未被重新应用升级。
3.1 滚动更新
deployment资源管理pod 滚动更新的过程 版本需要升级，当前的v1版本的pod数量为3 控制器会先杀死一个pod，并且通过deployment控制器会操作rs控制器生产v2版本的pod。保障服务的不间断访问，逐步迁移所有pod为v2版本的pod
3.2 回滚降级
在 deployment 滚动升级完成后 原v1版本的 rs控制器的资源清单不会被删除 而是存储在etcd数据库中
逐步删除v2版本的rs控制器的pod 通过v1版本的rs 生成v1版本的pod， 再次更新v2.1 和上面滚动升级同理，创建v2.1的rs控制器 逐步替换v1 的pod
3.3 Kubectl create、apply、replace 的区别
create 是一种命令式的表达方式 根据资源清单创建资源 apply 是一种声明式的表达方式 根据资源清单更新不同 replace 是一种命令式的表达方式 根据资源清单删除原有资源重新创建资源
# create 方式创建资源
[root@k8s-master ~]# cat deployment-1.yml
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: deployment-demo
name: deployment-demo
spec:
selector:
matchLabels:
app: deployment-demo
template:
metadata:
labels:
app: deployment-demo
spec:
containers:
- image: wangyanglinux/myapp:v1.0
name: deployment-demo-container
[root@k8s-master ~]# kubectl create -f deployment-1.yml
deployment.apps/deployment-demo created
[root@k8s-master ~]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
deployment-demo-6995c75668-d7t5j 1/1 Running 0 29s 10.244.85.247 k8s-node01 <none> <none>
[root@k8s-master ~]# curl 10.244.85.247
www.xinxianghf.com | hello MyAPP | version v1.0
## 修改资源清单
[root@k8s-master ~]# sed -i 's/myapp:v1.0/myapp:v2.0/' deployment-1.yml
[root@k8s-master ~]# kubectl create -f deployment-1.yml # 是无法在通过create创建资源的
Error from server (AlreadyExists): error when creating "deployment-1.yml": deployments.apps "deployment-demo" already exists
# apply 方式创建资源 [root@k8s-master ~]# kubectl delete -f deployment-1.yml deployment.apps "deployment-demo" deleted [root@k8s-master ~]# kubectl get pod -o wide No resources found in default namespace. [root@k8s-master ~]# kubectl apply -f deployment-1.yml deployment.apps/deployment-demo created [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-625rx 1/1 Running 0 67s 10.244.85.248 k8s-node01 <none> <none> [root@k8s-master ~]# curl 10.244.85.248 www.xinxianghf.com | hello MyAPP | version v2.0 [root@k8s-master ~]# sed -i 's/myapp:v2.0/myapp:v3.0/' deployment-1.yml [root@k8s-master ~]# kubectl apply -f deployment-1.yml # 可以更新现有资源与清单不同的地方 deployment.apps/deployment-demo configured [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-7dbccb74d6-brnjv 1/1 Running 0 2m39s 10.244.58.240 k8s-node02 <none> <none> [root@k8s-master ~]# curl 10.244.58.240 # 2.0 更新至3.0 www.xinxianghf.com | hello MyAPP | version v3.0
# replace 方式创建资源 [root@k8s-master ~]# sed -i 's/myapp:v3.0/myapp:v2.0/' deployment-1.yml [root@k8s-master ~]# kubectl replace -f deployment-1.yml # 直接删除原有v3版本pod重新创建2.0的pod deployment.apps/deployment-demo replaced [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-gwntz 0/1 ContainerCreating 0 3s <none> k8s-node01 <none> <none> deployment-demo-7dbccb74d6-brnjv 1/1 Running 0 4m26s 10.244.58.240 k8s-node02 <none> <none> [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-gwntz 1/1 Running 0 25s 10.244.85.249 k8s-node01 <none> <none> [root@k8s-master ~]# curl 10.244.85.249 www.xinxianghf.com | hello MyAPP | version v2.0
3.4 Deployment 常用命令
# --record 参数可以记录一些命令。 可以放方便查看每次reversion的变化 [root@k8s-master ~]# kubectl create -f deployment.yml --record # scale 修改名为nginx-deployment的deployment资源控制器副本数量为10 [root@k8s-master ~]# kubectl scale deployment nginx-deployment --replicas=10 # autoscale 自动伸缩扩展pod副本数量，cpu使用率在80% 那么会扩展pod，设置min最小值防止低峰pod回收后突然高峰顶爆当前这台设备，设置最大值防止被攻击或代码问题导致一直扩展pod 非真实用户访问造成资源浪费 [root@k8s-master ~]# kubectl autoscale deployment nginx-deployment --min=10 --max=15 --cpu-percent=80 # 设置名为deployment-demo的deployment资源控制器的容器名为deployment-demo-container的镜像为wangyanglinux/myapp:v1.0 [root@k8s-master ~]# kubectl set image deployment/deployment-demo deployment-demo-container=wangyanglinux/myapp:v1.0 deployment.apps/deployment-demo image updated
3.5 滚动更新实验
deployment-demo-6995c75668-qn8jv 0/1 ContainerCreating 0 2s [root@k8s-master ~]# cat deployment.yml apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: replicas: 5 selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v1.0 name: deployment-demo-container [root@k8s-master ~]# kubectl apply -f deployment.yml deployment.apps/deployment-demo create [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-6995c75668-25gxf 1/1 Running 0 20s deployment-demo-6995c75668-2tlhq 1/1 Running 0 20s deployment-demo-6995c75668-jsz5z 1/1 Running 0 20s deployment-demo-6995c75668-kdrx8 1/1 Running 0 73s deployment-demo-6995c75668-qn8jv 1/1 Running 0 20s [root@k8s-master ~]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS deployment-demo-6995c75668-25gxf 1/1 Running 0 113s app=deployment-demo,pod-template-hash=6995c75668 deployment-demo-6995c75668-2tlhq 1/1 Running 0 113s app=deployment-demo,pod-template-hash=6995c75668 deployment-demo-6995c75668-jsz5z 1/1 Running 0 113s app=deployment-demo,pod-template-hash=6995c75668 deployment-demo-6995c75668-kdrx8 1/1 Running 0 2m46s app=deployment-demo,pod-template-hash=6995c75668 deployment-demo-6995c75668-qn8jv 1/1 Running 0 113s app=deployment-demo,pod-template-hash=6995c75668 [root@k8s-master ~]# kubectl create svc clusterip deployment-demo --tcp=80:80 # 创建deployment控制器根据标签deployment-demo关系负载均衡集群 service/deployment-demo created [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE deployment-demo ClusterIP 10.2.163.19 <none> 80/TCP 5s kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 3d3h [root@k8s-master ~]# curl 10.2.163.19 # 访问负载均衡集群IP www.xinxianghf.com | hello MyAPP | version v1.0 # 再开一个终端循环访问集群的ip [root@k8s-master ~]# while true; do curl 10.2.163.19 && sleep 1 ; done # 循环访问负载集群的ip www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 # 回到原终端更新pod的容器镜像版本 [root@k8s-master ~]# kubectl set image deployment/deployment-demo deployment-demo-container=wangyanglinux/myapp:v2.0 deployment.apps/deployment-demo image updated [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-2xxnq 0/1 ContainerCreating 0 4s <none> k8s-node01 <none> <none> deployment-demo-6465d4c5c9-6gdff 1/1 Running 0 18s 10.244.85.254 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-6zspk 1/1 Running 0 17s 10.244.85.255 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-n4fb5 1/1 Running 0 17s 10.244.58.243 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-w748p 0/1 ContainerCreating 0 6s <none> k8s-node02 <none> <none> deployment-demo-6995c75668-25gxf 1/1 Terminating 0 8m44s 10.244.58.242 k8s-node02 <none> <none> deployment-demo-6995c75668-2tlhq 1/1 Terminating 0 8m44s 10.244.85.252 k8s-node01 <none> <none> deployment-demo-6995c75668-kdrx8 1/1 Terminating 0 9m37s 10.244.85.251 k8s-node01 <none> <none> deployment-demo-6995c75668-qn8jv 1/1 Running 0 8m44s 10.244.58.241 k8s-node02 <none> <none> [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES deployment-demo-6465d4c5c9-2xxnq 1/1 Running 0 27s 10.244.85.192 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-6gdff 1/1 Running 0 41s 10.244.85.254 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-6zspk 1/1 Running 0 40s 10.244.85.255 k8s-node01 <none> <none> deployment-demo-6465d4c5c9-n4fb5 1/1 Running 0 40s 10.244.58.243 k8s-node02 <none> <none> deployment-demo-6465d4c5c9-w748p 1/1 Running 0 29s 10.244.58.244 k8s-node02 <none> <none> # 回到循环访问集群的ip的终端， 可以发现v1版本逐渐与v2版本交替访问，直到完全被v2版本取代，这就是滚动升级的过程 [root@k8s-master ~]# while true; do curl 10.2.163.19 && sleep 1 ; done www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v1.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0 www.xinxianghf.com | hello MyAPP | version v2.0
3.6 Deployment-金丝雀部署
# 续接3.5滚动更新的实验，先删除deployment控制器创建的pod，保留svc [root@k8s-master ~]# kubectl delete -f deployment.yml [root@k8s-master ~]# kubectl get pod No resources found in default namespace. [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE deployment-demo ClusterIP 10.2.163.19 <none> 80/TCP 20h kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 3d23h
[root@k8s-master ~]# cat deployment.yml apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: replicas: 10 # 这里演示10个副本pod selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v1.0 name: deployment-demo-container
[root@k8s-master ~]# kubectl apply -f deployment.yml # 创建deploy控制器
deployment.apps/deployment-demo created
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6995c75668-crh9g 1/1 Running 0 80s
deployment-demo-6995c75668-fk2fj 1/1 Running 0 80s
deployment-demo-6995c75668-glzph 1/1 Running 0 80s
deployment-demo-6995c75668-ln92q 1/1 Running 0 81s
deployment-demo-6995c75668-lq76f 1/1 Running 0 81s
deployment-demo-6995c75668-m4nmn 1/1 Running 0 80s
deployment-demo-6995c75668-q5969 1/1 Running 0 80s
deployment-demo-6995c75668-rfpl5 1/1 Running 0 81s
deployment-demo-6995c75668-th6jv 1/1 Running 0 80s
deployment-demo-6995c75668-z2tl5 1/1 Running 0 80s
[root@k8s-master ~]# kubectl get svc
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
deployment-demo ClusterIP 10.2.163.19 <none> 80/TCP 20h
kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 4d
[root@k8s-master ~]# kubectl get deployment deployment-demo -o yaml # 把deployment-demo控制器输出为资源清单
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment","metadata":{"annotations":{},"labels":{"app":"deployment-demo"},"name":"deployment-demo","namespace":"default"},"spec":{"replicas":10,"selector":{"matchLabels":{"app":"deployment-demo"}},"template":{"metadata":{"labels":{"app":"deployment-demo"}},"spec":{"containers":[{"image":"wangyanglinux/myapp:v1.0","name":"deployment-demo-container"}]}}}}
creationTimestamp: "2024-12-21T11:13:06Z"
generation: 1
labels:
app: deployment-demo
name: deployment-demo
namespace: default
resourceVersion: "516213"
uid: 6a0e236a-c726-46cd-a5f0-4a6411d51297
spec:
progressDeadlineSeconds: 600
replicas: 10
revisionHistoryLimit: 10
selector:
matchLabels:
app: deployment-demo
strategy:
rollingUpdate: # 更新模式为滚动更新
maxSurge: 25% # 滚动更新允许创建超过pod副本的数量 这里是总副本数量的25%
maxUnavailable: 25% # 更新时pod最多不可用副本的数量 这里是总副本数量的25%
type: RollingUpdate
template:
metadata:
creationTimestamp: null
labels:
app: deployment-demo
spec:
containers:
- image: wangyanglinux/myapp:v1.0
imagePullPolicy: IfNotPresent
name: deployment-demo-container
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
dnsPolicy: ClusterFirst
restartPolicy: Always
schedulerName: default-scheduler
securityContext: {}
terminationGracePeriodSeconds: 30
status:
availableReplicas: 10
conditions:
- lastTransitionTime: "2024-12-21T11:14:01Z"
lastUpdateTime: "2024-12-21T11:14:01Z"
message: Deployment has minimum availability.
reason: MinimumReplicasAvailable
status: "True"
type: Available
- lastTransitionTime: "2024-12-21T11:13:06Z"
lastUpdateTime: "2024-12-21T11:14:06Z"
message: ReplicaSet "deployment-demo-6995c75668" has successfully progressed.
reason: NewReplicaSetAvailable
status: "True"
type: Progressing
observedGeneration: 1
readyReplicas: 10
replicas: 10
updatedReplicas: 10
# 配置deployment-demo控制器的滚动更新策略为创建新pod版本的数量为1，不可用pod副本的数量为0
[root@k8s-master ~]# kubectl patch deployment deployment-demo -p '{"spec":{"strategy":{"rollingUpdate":{"maxSurge":1,"maxUnavailable":0}}}}'
deployment.apps/deployment-demo patched
[root@k8s-master ~]# kubectl get deployment deployment-demo -o yaml
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment","metadata":{"annotations":{},"labels":{"app":"deployment-demo"},"name":"deployment-demo","namespace":"default"},"spec":{"replicas":10,"selector":{"matchLabels":{"app":"deployment-demo"}},"template":{"metadata":{"labels":{"app":"deployment-demo"}},"spec":{"containers":[{"image":"wangyanglinux/myapp:v1.0","name":"deployment-demo-container"}]}}}}
creationTimestamp: "2024-12-21T11:13:06Z"
generation: 2
labels:
app: deployment-demo
name: deployment-demo
namespace: default
resourceVersion: "517361"
uid: 6a0e236a-c726-46cd-a5f0-4a6411d51297
spec:
progressDeadlineSeconds: 600
replicas: 10
revisionHistoryLimit: 10
selector:
matchLabels:
app: deployment-demo
strategy:
rollingUpdate:
maxSurge: 1 # 滚动更新允许创建超过pod副本的数量1 那么滚动升级是会出现11个容器
maxUnavailable: 0 # 不可用pod副本的数量为0 也就是更新的过程中，新pod可以访问后才会删除一个旧pod，而不是默认删除pod副本数量的25%
type: RollingUpdate
template:
metadata:
creationTimestamp: null
labels:
app: deployment-demo
spec:
containers:
- image: wangyanglinux/myapp:v1.0
imagePullPolicy: IfNotPresent
name: deployment-demo-container
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
dnsPolicy: ClusterFirst
restartPolicy: Always
schedulerName: default-scheduler
securityContext: {}
terminationGracePeriodSeconds: 30
status:
availableReplicas: 10
conditions:
- lastTransitionTime: "2024-12-21T11:14:01Z"
lastUpdateTime: "2024-12-21T11:14:01Z"
message: Deployment has minimum availability.
reason: MinimumReplicasAvailable
status: "True"
type: Available
- lastTransitionTime: "2024-12-21T11:13:06Z"
lastUpdateTime: "2024-12-21T11:14:06Z"
message: ReplicaSet "deployment-demo-6995c75668" has successfully progressed.
reason: NewReplicaSetAvailable
status: "True"
type: Progressing
observedGeneration: 2
readyReplicas: 10
replicas: 10
updatedReplicas: 10
# kubectl patch deployment deployment-demo -p '{"spec":{"template":{"spec":{"containers":[{"name":"deployment-demo-container","image":"wangyanglinux/myapp:v2.0"}]}}}}' # 触发滚动更新把镜像改为2.0版本
# kubectl rollout pause deployment deployment-demo 暂停名为deployment-demo的deploy控制器滚动更新
[root@k8s-master ~]# kubectl patch deployment deployment-demo -p '{"spec":{"template":{"spec":{"containers":[{"name":"deployment-demo-container","image":"wangyanglinux/myapp:v2.0"}]}}}}' && kubectl rollout pause deployment deployment-demo
deployment.apps/deployment-demo patched
deployment.apps/deployment-demo paused
[root@k8s-master ~]# kubectl get pod # 查看pod 发现滚动更新创建了上面定义的pod数量 为1个pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6465d4c5c9-jsnsl 0/1 ContainerCreating 0 7s # 也就是这个最新创建的
deployment-demo-6995c75668-crh9g 1/1 Running 0 22m
deployment-demo-6995c75668-fk2fj 1/1 Running 0 22m
deployment-demo-6995c75668-glzph 1/1 Running 0 22m
deployment-demo-6995c75668-ln92q 1/1 Running 0 22m
deployment-demo-6995c75668-lq76f 1/1 Running 0 22m
deployment-demo-6995c75668-m4nmn 1/1 Running 0 22m
deployment-demo-6995c75668-q5969 1/1 Running 0 22m
deployment-demo-6995c75668-rfpl5 1/1 Running 0 22m
deployment-demo-6995c75668-th6jv 1/1 Running 0 22m
deployment-demo-6995c75668-z2tl5 1/1 Running 0 22m
[root@k8s-master ~]# kubectl get deploy # 查看deploy控制器发现 容器数量为11，更新1个pod
NAME READY UP-TO-DATE AVAILABLE AGE
deployment-demo 11/10 1 11 25m
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6465d4c5c9-jsnsl 1/1 Running 0 2m25s
deployment-demo-6995c75668-crh9g 1/1 Running 0 25m
deployment-demo-6995c75668-fk2fj 1/1 Running 0 25m
deployment-demo-6995c75668-glzph 1/1 Running 0 25m
deployment-demo-6995c75668-ln92q 1/1 Running 0 25m
deployment-demo-6995c75668-lq76f 1/1 Running 0 25m
deployment-demo-6995c75668-m4nmn 1/1 Running 0 25m
deployment-demo-6995c75668-q5969 1/1 Running 0 25m
deployment-demo-6995c75668-rfpl5 1/1 Running 0 25m
deployment-demo-6995c75668-th6jv 1/1 Running 0 25m
deployment-demo-6995c75668-z2tl5 1/1 Running 0 25m
# 打开新终端，测试访问svc 负载均衡 11个pod 负载均衡的方式访问
[root@k8s-master ~]# while true ; do curl 10.2.163.19 && sleep 1 ; done
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
[root@k8s-master ~]# kubectl rollout resume deployment deployment-demo # 继续进行滚动升级
deployment.apps/deployment-demo resumed
[root@k8s-master ~]# kubectl get pod # 再次查看pod，可以发现通过设置以后的pod runing的最低数量为10，并且同时进行更新1个pod，删除一个pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6465d4c5c9-jsnsl 1/1 Running 0 8m
deployment-demo-6465d4c5c9-sr8p7 0/1 Pending 0 2s
deployment-demo-6995c75668-crh9g 1/1 Running 0 30m
deployment-demo-6995c75668-fk2fj 1/1 Running 0 30m
deployment-demo-6995c75668-glzph 1/1 Running 0 30m
deployment-demo-6995c75668-ln92q 1/1 Terminating 0 30m
deployment-demo-6995c75668-lq76f 1/1 Running 0 30m
deployment-demo-6995c75668-m4nmn 1/1 Running 0 30m
deployment-demo-6995c75668-q5969 1/1 Running 0 30m
deployment-demo-6995c75668-rfpl5 1/1 Running 0 30m
deployment-demo-6995c75668-th6jv 1/1 Running 0 30m
deployment-demo-6995c75668-z2tl5 1/1 Running 0 30m
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6465d4c5c9-jsnsl 1/1 Running 0 8m4s
deployment-demo-6465d4c5c9-sr8p7 0/1 ContainerCreating 0 6s
deployment-demo-6995c75668-crh9g 1/1 Running 0 30m
deployment-demo-6995c75668-fk2fj 1/1 Running 0 30m
deployment-demo-6995c75668-glzph 1/1 Running 0 30m
deployment-demo-6995c75668-ln92q 1/1 Terminating 0 30m
deployment-demo-6995c75668-lq76f 1/1 Running 0 30m
deployment-demo-6995c75668-m4nmn 1/1 Running 0 30m
deployment-demo-6995c75668-q5969 1/1 Running 0 30m
deployment-demo-6995c75668-rfpl5 1/1 Running 0 30m
deployment-demo-6995c75668-th6jv 1/1 Running 0 30m
deployment-demo-6995c75668-z2tl5 1/1 Running 0 30m
# 滚动降级，也就是回滚之前的版本
[root@k8s-master ~]# kubectl rollout undo deployment deployment-demo # 回滚
deployment.apps/deployment-demo rolled back
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6465d4c5c9-bxmwn 1/1 Running 0 8m52s
deployment-demo-6465d4c5c9-dt6xg 1/1 Running 0 8m17s
deployment-demo-6465d4c5c9-hx252 1/1 Running 0 10m
deployment-demo-6465d4c5c9-jsnsl 1/1 Running 0 18m
deployment-demo-6465d4c5c9-l4kwg 1/1 Running 0 9m42s
deployment-demo-6465d4c5c9-mhm6s 1/1 Running 0 10m
deployment-demo-6465d4c5c9-nzwgb 1/1 Running 0 9m8s
deployment-demo-6465d4c5c9-rn8w2 1/1 Running 0 8m35s
deployment-demo-6465d4c5c9-sr8p7 1/1 Running 0 10m
deployment-demo-6465d4c5c9-tgtqh 1/1 Running 0 9m25s
deployment-demo-6995c75668-zhb75 0/1 ContainerCreating 0 9s # 可以看到回滚到v1的pod被创建了
# 验证查看pod的版本
[root@k8s-master ~]# while true ; do curl 10.2.163.19 && sleep 1 ; done
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v2.0
www.xinxianghf.com | hello MyAPP | version v2.0
# 等待片刻 回滚完毕查看pod是否全部回滚完成running的状态
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
deployment-demo-6995c75668-25glq 1/1 Running 0 2m22s
deployment-demo-6995c75668-2tfx5 1/1 Running 0 105s
deployment-demo-6995c75668-62m2m 1/1 Running 0 73s
deployment-demo-6995c75668-bmz4m 1/1 Running 0 2m4s
deployment-demo-6995c75668-mcnq5 1/1 Running 0 41s
deployment-demo-6995c75668-tjm65 1/1 Running 0 57s
deployment-demo-6995c75668-v9h8m 1/1 Running 0 90s
deployment-demo-6995c75668-vggmm 1/1 Running 0 2m39s
deployment-demo-6995c75668-x4h8p 1/1 Running 0 23s
deployment-demo-6995c75668-zhb75 1/1 Running 0 2m50s
# 验证访问回滚后的pod版本信息都是v1
[root@k8s-master ~]# while true ; do curl 10.2.163.19 && sleep 1 ; done
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
www.xinxianghf.com | hello MyAPP | version v1.0
# deployment实现滚动升级也是通过控制rs控制器进行的升级回滚，可以看到这里被deployment控制器创建出2个rs控制器
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
deployment-demo-6465d4c5c9 0 0 0 23m # 根据上面的实验得知 这个pod为0的就是v2版
deployment-demo-6995c75668 10 10 10 46m # 根据上面的实验得知 这个pod为10的就是v1版
生产环境中，滚动升级的幅度不建议超过25%，也就是同时创建的pod数量不超过当前副本数量的25%以免造成业务的不稳定性。 限制不可用pod副本的数量为0 可以有效的避免同时删除大量pod线上用户大批量的连接中断给用户造成不好的体验。 可以根据实际的生产环境设置滚动升级创建pod副本的数量以及不可用容器的数量，
3.7 清理策略
Deployment 创建资源清单后，进行版本回退以及版本更新时都会创建rs控制器并存储在etcd数据库中
# 上面章节中提到回滚以及版本更新大多数都是基于命令修改镜像版本做升级，这里演示使用修改资源清单文件做升级 [root@k8s-master ~]# cat deployment.yml apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: replicas: 2 selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v1.0 name: deployment-demo-container [root@k8s-master ~]# kubectl apply -f deployment.yml deployment.apps/deployment-demo created [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-6995c75668-58mrm 1/1 Running 0 13s deployment-demo-6995c75668-pqzzb 1/1 Running 0 12s [root@k8s-master ~]# kubectl get rs # deployment生成的rs控制器 NAME DESIRED CURRENT READY AGE deployment-demo-6995c75668 2 2 2 65s # 修改镜像版本 [root@k8s-master ~]# sed -i 's/v1.0/v2.0/' deployment.yml # 镜像升级为2.0 [root@k8s-master ~]# cat deployment.yml apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: replicas: 2 selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v2.0 # 这里已经修改为2.0 name: deployment-demo-container # 升级版本，获取历史版本信息 [root@k8s-master ~]# kubectl apply -f deployment.yml deployment.apps/deployment-demo configured [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-6465d4c5c9-2lnlc 1/1 Running 0 22s deployment-demo-6465d4c5c9-688kr 1/1 Running 0 15s [root@k8s-master ~]# kubectl get rs NAME DESIRED CURRENT READY AGE deployment-demo-6465d4c5c9 2 2 2 108s # 这个就是v2.0的rs了 deployment-demo-6995c75668 0 0 0 5m49s # 这个是v1.0的rs 也就是旧版本 [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE deployment-demo ClusterIP 10.2.163.19 <none> 80/TCP 3d12h kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 6d16h [root@k8s-master ~]# curl 10.2.163.19 # 访问验证 www.xinxianghf.com | hello MyAPP | version v2.0 # 不保留版本信息到etcd数据库中 # 删除历史所有deploy控制器 [root@k8s-master ~]# kubectl delete deployment --all deployment.apps "deployment-demo" deleted [root@k8s-master ~]# kubectl get rs No resources found in default namespace. [root@k8s-master ~]# kubectl get pod No resources found in default namespace. # revisionHistoryLimit: 0 不保留历史版本 [root@k8s-master ~]# cat deployment.yml apiVersion: apps/v1 kind: Deployment metadata: labels: app: deployment-demo name: deployment-demo spec: revisionHistoryLimit: 0 # 添加到控制器中 replicas: 2 selector: matchLabels: app: deployment-demo template: metadata: labels: app: deployment-demo spec: containers: - image: wangyanglinux/myapp:v2.0 name: deployment-demo-container [root@k8s-master ~]# kubectl apply -f deployment.yml # 创建不保留历史版本的控制器 deployment.apps/deployment-demo created [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-6465d4c5c9-8zghd 1/1 Running 0 25s deployment-demo-6465d4c5c9-qw7x6 1/1 Running 0 25s [root@k8s-master ~]# sed -i 's/v2.0/v3.0/' deployment.yml # 修改版本 [root@k8s-master ~]# kubectl apply -f deployment.yml deployment.apps/deployment-demo configured [root@k8s-master ~]# kubectl get pod NAME READY STATUS RESTARTS AGE deployment-demo-7dbccb74d6-trbws 1/1 Running 0 62s deployment-demo-7dbccb74d6-xshg9 1/1 Running 0 54s [root@k8s-master ~]# curl 10.2.163.19 www.xinxianghf.com | hello MyAPP | version v3.0 [root@k8s-master ~]# kubectl get rs # 这里可以看到是没有历史的rs控制器 NAME DESIRED CURRENT READY AGE deployment-demo-7dbccb74d6 2 2 2 19s # 这样的方式的好处是通过资源清单来做版本的回滚以及更新，rs控制器不会存储到etcd数据库中，在做版本变更时应该先备份原版的资源清单在做修改并添加变更的说明
# 删除创建的控制器 防止影响下章节 [root@k8s-master ~]# kubectl delete deployment --all deployment.apps "deployment-demo" deleted [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE deployment-demo ClusterIP 10.2.163.19 <none> 80/TCP 3d12h kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 6d16h [root@k8s-master ~]# kubectl delete svc deployment-demo service "deployment-demo" deleted
四、DaemonSet 控制器
# 创建DaemonSet控制器资源清单文件 [root@k8s-master ~]# cat >> Daemonset.yml <<<EOF apiVersion: apps/v1 kind: DaemonSet metadata: name: daemonset-demo labels: app: daemonset-demo spec: selector: matchLabels: name: daemonset-demo template: metadata: labels: name: daemonset-demo spec: containers: - name: daemonset-demo-container image: wangyanglinux/myapp:v1.0 EOF [root@k8s-master ~]# kubectl apply -f Daemonset.yml daemonset.apps/daemonset-demo created [root@k8s-master ~]# kubectl get pod -o wide # 查看pod详细信息，可以发现node01，node02各一个，而资源清单中并没有写副本数量， NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES daemonset-demo-pc97q 1/1 Running 0 12s 10.244.58.196 k8s-node02 <none> <none> daemonset-demo-t8blw 1/1 Running 0 12s 10.244.85.221 k8s-node01 <none> <none> [root@k8s-master ~]# kubectl delete daemonset daemonset-demo # 删除
七、Service控制器
修改kube-proxy 工作模式
[root@k8s-master ~]# kubectl edit configmap kube-proxy -n kube-system # 修改 mode: "ipvs" [root@k8s-master ~]# kubectl get pod -l k8s-app=kube-proxy -A # 查询 kube-proxy 标签的pod NAMESPACE NAME READY STATUS RESTARTS AGE kube-system kube-proxy-86hs7 1/1 Running 3 (27h ago) 6d22h kube-system kube-proxy-mhzph 1/1 Running 1 (27h ago) 6d22h kube-system kube-proxy-vjfsw 1/1 Running 6 (27h ago) 6d22h # 修改完后需要删除程序，自动重建 [root@k8s-master ~]# kubectl delete pod -n kube-system -l k8s-app=kube-proxy pod "kube-proxy-86hs7" deleted pod "kube-proxy-mhzph" deleted pod "kube-proxy-vjfsw" deleted # 查看是否重建成功 Running说明成功 [root@k8s-master ~]# kubectl get pod -l k8s-app=kube-proxy -A NAMESPACE NAME READY STATUS RESTARTS AGE kube-system kube-proxy-c6fht 1/1 Running 0 2m49s kube-system kube-proxy-gs7gw 1/1 Running 0 2m48s kube-system kube-proxy-jcs4b 1/1 Running 0 2m49s
# 就绪探测 readinessProbe: httpGet: port: 80 # 探测端口 path: /index1.html # 探测是否有这个文件，有则就绪，没有就非就绪 initialDelaySeconds: 1 periodSeconds: 3 ------------- apiVersion: v1 kind: Service metadata: name: myapp-clusterip namespace: default spec: type: ClusterIP # ClusterIP类型 selector: app: myapp release: stabel svc: clusterip ports: - name: http port: 80 # 负载均衡的访问端口 targetPort: 80 # 后端服务端口
7.1 Clusterip类型
[root@k8s-master ~]# cat >> myapp-clusterip-deploy.yaml <EOF apiVersion: apps/v1 kind: Deployment metadata: name: myapp-clusterip-deploy namespace: default spec: replicas: 3 selector: matchLabels: app: myapp release: stabel svc: clusterip template: metadata: labels: app: myapp release: stabel env: test svc: clusterip spec: containers: - name: myapp-container image: wangyanglinux/myapp:v1.0 imagePullPolicy: IfNotPresent ports: - name: http containerPort: 80 readinessProbe: httpGet: port: 80 path: /index1.html initialDelaySeconds: 1 periodSeconds: 3 EOF [root@k8s-master ~]# kubectl apply -f myapp-clusterip-deploy.yaml deployment.apps/myapp-clusterip-deploy created [root@k8s-master ~]# kubectl get pod # 可以看就绪状态是非就绪 NAME READY STATUS RESTARTS AGE myapp-clusterip-deploy-5c9cc9b64-2ssbv 0/1 Running 0 41s myapp-clusterip-deploy-5c9cc9b64-pf6b7 0/1 Running 0 41s myapp-clusterip-deploy-5c9cc9b64-pg4ww 0/1 Running 0 41s # 创建svc控制器资源 [root@k8s-master ~]# cat >> myapp-service.yaml <<EOF apiVersion: v1 kind: Service metadata: name: myapp-clusterip namespace: default spec: type: ClusterIP selector: app: myapp release: stabel svc: clusterip ports: - name: http port: 80 targetPort: 80 EOF [root@k8s-master ~]# kubectl apply -f myapp-service.yaml service/myapp-clusterip created [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 6d21h myapp-clusterip ClusterIP 10.0.192.93 <none> 80/TCP 7s [root@k8s-master ~]# curl 10.0.192.93 # 访问这个svc ip发现无法访问 curl: (7) Failed to connect to 10.0.192.93 port 80: Connection refused # 查看pod标签以及svc标签，发现pod的标签是svc标签的子集 [root@k8s-master ~]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS myapp-clusterip-deploy-5c9cc9b64-2ssbv 0/1 Running 0 8m24s app=myapp,env=test,pod-template-hash=5c9cc9b64,release=stabel,svc=clusterip myapp-clusterip-deploy-5c9cc9b64-pf6b7 0/1 Running 0 8m24s app=myapp,env=test,pod-template-hash=5c9cc9b64,release=stabel,svc=clusterip myapp-clusterip-deploy-5c9cc9b64-pg4ww 0/1 Running 0 8m24s app=myapp,env=test,pod-template-hash=5c9cc9b64,release=stabel,svc=clusterip [root@k8s-master ~]# kubectl get svc -o wide NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE SELECTOR kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 6d21h <none> myapp-clusterip ClusterIP 10.0.192.93 <none> 80/TCP 3m20s app=myapp,release=stabel,svc=clusterip # 创建就绪检测文件 [root@k8s-master ~]# kubectl exec -it myapp-clusterip-deploy-5c9cc9b64-2ssbv -- /bin/bash # 进入pod中 myapp-clusterip-deploy-5c9cc9b64-2ssbv:/# cd /usr/local/nginx/html/ # 把当前时间重定向到index1.html文件中 myapp-clusterip-deploy-5c9cc9b64-2ssbv:/usr/local/nginx/html# date > index1.html [root@k8s-master ~]# kubectl get pod # myapp-clusterip-deploy-5c9cc9b64-2ssbv 这个pod已经就绪 NAME READY STATUS RESTARTS AGE myapp-clusterip-deploy-5c9cc9b64-2ssbv 1/1 Running 0 11m myapp-clusterip-deploy-5c9cc9b64-pf6b7 0/1 Running 0 11m myapp-clusterip-deploy-5c9cc9b64-pg4ww 0/1 Running 0 11m [root@k8s-master ~]# curl 10.0.192.93 # 发现也可以访问 www.xinxianghf.com | hello MyAPP | version v1.0 [root@k8s-master ~]# curl 10.0.192.93/hostname.html # 访问打印pod名的问题，发现只能访问到就绪pod myapp-clusterip-deploy-5c9cc9b64-2ssbv # 在创建一个就绪文件 [root@k8s-master ~]# kubectl exec -it myapp-clusterip-deploy-5c9cc9b64-pf6b7 -- /bin/bash myapp-clusterip-deploy-5c9cc9b64-pf6b7:/# date > /usr/local/nginx/html/index1.html myapp-clusterip-deploy-5c9cc9b64-pf6b7:/# exit exit [root@k8s-master ~]# kubectl get pod # 发现已经两个pod就绪 NAME READY STATUS RESTARTS AGE myapp-clusterip-deploy-5c9cc9b64-2ssbv 1/1 Running 0 19m myapp-clusterip-deploy-5c9cc9b64-pf6b7 1/1 Running 0 19m myapp-clusterip-deploy-5c9cc9b64-pg4ww 0/1 Running 0 19m [root@k8s-master ~]# curl 10.0.192.93/hostname.html myapp-clusterip-deploy-5c9cc9b64-pf6b7 [root@k8s-master ~]# curl 10.0.192.93/hostname.html myapp-clusterip-deploy-5c9cc9b64-2ssbv # 那么就证实了 svc 关联到pod实现负载均衡就必须满足2个条件。 # 1、负载的pod标签必须是svc控制器的子集匹配，或者全集 # 2、pod 必须处于就绪状态 svc才会进行提供负载均衡访问 # 查看ipvs转发规则 [root@k8s-master ~]# ipvsadm -Ln IP Virtual Server version 1.2.1 (size=4096) Prot LocalAddress:Port Scheduler Flags -> RemoteAddress:Port Forward Weight ActiveConn InActConn TCP 10.0.0.1:443 rr -> 192.168.68.160:6443 Masq 1 0 0 TCP 10.0.0.10:53 rr -> 10.244.58.204:53 Masq 1 0 0 -> 10.244.58.209:53 Masq 1 0 0 TCP 10.0.0.10:9153 rr -> 10.244.58.204:9153 Masq 1 0 0 -> 10.244.58.209:9153 Masq 1 0 0 TCP 10.0.192.93:80 rr # 这个就是创建的scv的规则 rr 轮询算法 -> 10.244.58.197:80 Masq 1 0 0 # 后端服务器 -> 10.244.85.224:80 Masq 1 0 0 # 后端服务器 TCP 10.6.2.65:5473 rr -> 192.168.68.161:5473 Masq 1 0 0 UDP 10.0.0.10:53 rr -> 10.244.58.204:53 Masq 1 0 0 -> 10.244.58.209:53 Masq 1 0 0 [root@k8s-master ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 6d22h myapp-clusterip ClusterIP 10.0.192.93 <none> 80/TCP 71m [root@k8s-master ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES myapp-clusterip-deploy-5c9cc9b64-2ssbv 1/1 Running 0 77m 10.244.58.197 k8s-node02 <none> <none> myapp-clusterip-deploy-5c9cc9b64-pf6b7 1/1 Running 0 77m 10.244.85.224 k8s-node01 <none> <none> myapp-clusterip-deploy-5c9cc9b64-pg4ww 0/1 Running 0 77m 10.244.85.223 k8s-node01 <none> <none>
控制器总结
控制器
Pod控制器
RC控制器
保障当前的Pod数量与期望一致
RS控制器
功能与RC控制器类似，但是多了标签选择运算的方式。k8s官方也推荐使用RS而不是RC
Deployment控制器
支持了声明式的表达
支持滚动更新和回滚
原理 deloyment > rs > pod
DaemonSet 控制器
保障每个节点有且只有一个pod，动态保障。
Kubectl create 、 apply、 replace
create
创建资源对象
通过基于资源清单文件的创建，但是如果此文件描述的资源对象存在，那么文件描述发生了改变再次提交也不会应用
apply
创建资源对象、修改资源对象
通过基于资源清单文件的创建，如果目标对象文件与本身运行的资源对象发生改变，那么会根据改变的 文件指定一一倾向修改目标的属性 （部分更新）
replace
创建资源对象、修改资源对象
通过基于资源清单文件的创建，如果目标对象文件本身运行的资源对象发生改变，那么会重建此对象（完全替换）
svc控制器
选中进行负载均衡pod的逻辑 1、pod是处于就绪状态。2、pod的标签是svc标签的子集（同一个命名空间下）
声明：本站所有文章，如无特殊说明或标注，均为本站原创发布。任何个人或组织，在未征得本站同意时，禁止复制、盗用、采集、发布本站内容到任何网站、书籍等各类媒体平台。如若本站内容侵犯了原著者的合法权益，可联系我们进行处理。
评论(0)