devbox
煮酒与君饮
这个屌丝很懒，什么也没留下！
关注作者
热门标签
热门文章
当前位置:   article > 正文

k8s 1.18.20版本部署_flannel-cni-plugin-v1.1.2.tar.gz

作者：煮酒与君饮 | 2024-08-09 11:02:31

赞

踩

flannel-cni-plugin-v1.1.2.tar.gz

身为k8s初学者，在掌握k8s理论知识的同时，也需要掌握一下实际部署k8s的过程，对于理论的学习起到一定的帮助作用。罗列了一下相关步骤，请各位参考：

一、环境准备

三台虚机：

操作系统： CentOS Linux release 7.9.2009 (Core)   

内核版本：3.10.0-1160.88.1.el7.x86_64

k8s-master01: 192.168.66.200   

k8s-master02: 192.168.66.201

k8s-node01:    192.168.66.250

本次安装k8s 1.18.20版本

二、服务器初始化配置

服务器初始化配置可以参考原先我写的1.15.0版本部署方案

K8s 1.15.0 版本部署安装_好好学习之乘风破浪的博客-CSDN博客

三、安装docker

  1. 3.1、安装指定的Docker版本
  2. yum list docker-ce.x86_64 --showduplicates | sort -r
  3. yum install -y containerd.io-1.6.18 docker-ce-23.0.1 docker-ce-cli-23.0.1
  4.  
  5. 3.2、创建docker daemon.json文件
  6. mkdir /etc/docker
  7. cat > /etc/docker/daemon.json <<EOF
  8. {
  9.   "registry-mirrors": ["https://docker.mirrors.ustc.edu.cn"],
  10.   "exec-opts": ["native.cgroupdriver=systemd"],
  11.   "log-driver": "json-file",
  12.   "log-opts": {
  13.     "max-size": "100m"
  14.   },
  15.   "storage-driver": "overlay2",
  16.   "storage-opts": [
  17.     "overlay2.override_kernel_check=true"
  18.   ]
  19. }
  20. EOF
  21.  
  22. 3.3 启动docker
  23. systemctl enable docker; systemctl start docker;systemctl status docker
  24.  
  25. 3.4 查看安装后的docker版本
  26. docker --version

四、安装软件包

  1. 4.1、安装软件
  2. yum install -y kubelet-1.18.20 kubeadm-1.18.20 kubectl-1.18.20 ipvsadm
  3.  
  4. 4.2、配置kubelet
  5. cat >  /etc/sysconfig/kubelet <<EOF
  6. KUBELET_CGROUP_ARGS="--cgroup-driver=systemd"
  7. KUBELET_EXTRA_ARGS="--fail-swap-on=false"
  8. EOF
  9.  
  10. 4.3、启动kubelet并设置开机启动
  11. systemctl enable kubelet && systemctl start kubelet
  12.  
  13. 4.4、检查版本
  14. kubeadm version
  15. 确认是所安装的版本

五、部署集群

以下操作在k8s-master01节点上操作

5.1 、生成初始化配置
kubeadm config print init-defaults > kubeadm.yaml

  1. 5.1 、生成初始化配置
  2. kubeadm config print init-defaults > kubeadm.yaml
  3.  
  4. [root@k8s-master01 ~]# cat  kubeadm.yaml
  5. apiVersion: kubeadm.k8s.io/v1beta2
  6. bootstrapTokens:
  7. - groups:
  8.   - system:bootstrappers:kubeadm:default-node-token
  9.   token: abcdef.0123456789abcdef
  10.   ttl: 24h0m0s
  11.   usages:
  12.   - signing
  13.   - authentication
  14. kind: InitConfiguration
  15. localAPIEndpoint:
  16.   advertiseAddress: 192.168.66.200    ####需要修改为k8s-master01的地址
  17.   bindPort: 6443
  18. nodeRegistration:
  19.   criSocket: /var/run/dockershim.sock
  20.   name: k8s-master01                 ####需要修改为k8s-master01的主机名
  21.   taints:
  22.   - effect: NoSchedule
  23.     key: node-role.kubernetes.io/master
  24. ---
  25. apiServer:
  26.   timeoutForControlPlane: 4m0s
  27. apiVersion: kubeadm.k8s.io/v1beta2
  28. certificatesDir: /etc/kubernetes/pki
  29. clusterName: kubernetes
  30. controllerManager: {}
  31. dns:
  32.   type: CoreDNS
  33. etcd:
  34.   local:
  35.     dataDir: /var/lib/etcd
  36. imageRepository: k8s.gcr.io  
  37. kind: ClusterConfiguration
  38. kubernetesVersion: v1.18.20   ####需要修改为v1.18.20
  39. networking:
  40.   dnsDomain: cluster.local
  41.   serviceSubnet: 10.96.0.0/12  ####保持默认
  42.   podSubnet: 10.2.0.0/16     ####新增
  43. scheduler: {}
  44.  
  45. ---                                             ####新增
  46. apiVersion: kubeproxy.config.k8s.io/v1alpha1    ####新增
  47. kind: KubeProxyConfiguration                    ####新增
  48. mode: ipvs                                      ####新增

5.2 、执行初始化配置
kubeadm init --config kubeadm.yaml
成功会显示如下信息

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

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 192.168.66.200:6443 --token abcdef.0123456789abcdef \
    --discovery-token-ca-cert-hash sha256:8ba059662766394bbc081324dfba0bc6a1360687f99e93a1b5c7c3a1e6d53097 

5.3 按照要求执行生成文件

  mkdir -p $HOME/.kube
  sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
  sudo chown $(id -u):$(id -g) $HOME/.kube/config

5.4 查看组件状态

kubectl get cs

该状态其实不影响整个集群使用，是因为默认的kube-scheduler.yaml和kube-controller-manager.yaml 中默认使用了10251和10252的端口。可以将对应的配置文件中- --port=0 给注释掉

 

  1. [root@k8s-master01 ~]# sed -i 's/- --port=0/#&/' /etc/kubernetes/manifests/kube-scheduler.yaml
  2. [root@k8s-master01 ~]# sed -i 's/- --port=0/#&/' /etc/kubernetes/manifests/kube-controller-manager.yaml
  3. [root@k8s-master01 ~]#  systemctl restart kubelet

 

 5.5、部署flannel组件

mirrors / coreos / flannel · GitCode

可以自行去github,找到 flannel/Documentation/kube-flannel.yml 路径，拷贝出代码，新建文件上传，然后修改为.yml

# vi kube-flannel.yml

将

  net-conf.json: |
    {
      "Network": "10.2.0.0/16",      ####修改为podsubnet地址段
      "Backend": {
        "Type": "vxlan"
      }
    }

（注：镜像路径可以不用修改）

  1. [root@k8s-master01 ~]# cat kube-flannel.yml 
  2. ---
  3. kind: Namespace
  4. apiVersion: v1
  5. metadata:
  6.   name: kube-flannel
  7.   labels:
  8.     pod-security.kubernetes.io/enforce: privileged
  9. ---
  10. kind: ClusterRole
  11. apiVersion: rbac.authorization.k8s.io/v1
  12. metadata:
  13.   name: flannel
  14. rules:
  15. - apiGroups:
  16.   - ""
  17.   resources:
  18.   - pods
  19.   verbs:
  20.   - get
  21. - apiGroups:
  22.   - ""
  23.   resources:
  24.   - nodes
  25.   verbs:
  26.   - get
  27.   - list
  28.   - watch
  29. - apiGroups:
  30.   - ""
  31.   resources:
  32.   - nodes/status
  33.   verbs:
  34.   - patch
  35. - apiGroups:
  36.   - "networking.k8s.io"
  37.   resources:
  38.   - clustercidrs
  39.   verbs:
  40.   - list
  41.   - watch
  42. ---
  43. kind: ClusterRoleBinding
  44. apiVersion: rbac.authorization.k8s.io/v1
  45. metadata:
  46.   name: flannel
  47. roleRef:
  48.   apiGroup: rbac.authorization.k8s.io
  49.   kind: ClusterRole
  50.   name: flannel
  51. subjects:
  52. - kind: ServiceAccount
  53.   name: flannel
  54.   namespace: kube-flannel
  55. ---
  56. apiVersion: v1
  57. kind: ServiceAccount
  58. metadata:
  59.   name: flannel
  60.   namespace: kube-flannel
  61. ---
  62. kind: ConfigMap
  63. apiVersion: v1
  64. metadata:
  65.   name: kube-flannel-cfg
  66.   namespace: kube-flannel
  67.   labels:
  68.     tier: node
  69.     app: flannel
  70. data:
  71.   cni-conf.json: |
  72.     {
  73.       "name": "cbr0",
  74.       "cniVersion": "0.3.1",
  75.       "plugins": [
  76.         {
  77.           "type": "flannel",
  78.           "delegate": {
  79.             "hairpinMode": true,
  80.             "isDefaultGateway": true
  81.           }
  82.         },
  83.         {
  84.           "type": "portmap",
  85.           "capabilities": {
  86.             "portMappings": true
  87.           }
  88.         }
  89.       ]
  90.     }
  91.   net-conf.json: |
  92.     {
  93.       "Network": "10.2.0.0/16",
  94.       "Backend": {
  95.         "Type": "vxlan"
  96.       }
  97.     }
  98. ---
  99. apiVersion: apps/v1
  100. kind: DaemonSet
  101. metadata:
  102.   name: kube-flannel-ds
  103.   namespace: kube-flannel
  104.   labels:
  105.     tier: node
  106.     app: flannel
  107. spec:
  108.   selector:
  109.     matchLabels:
  110.       app: flannel
  111.   template:
  112.     metadata:
  113.       labels:
  114.         tier: node
  115.         app: flannel
  116.     spec:
  117.       affinity:
  118.         nodeAffinity:
  119.           requiredDuringSchedulingIgnoredDuringExecution:
  120.             nodeSelectorTerms:
  121.             - matchExpressions:
  122.               - key: kubernetes.io/os
  123.                 operator: In
  124.                 values:
  125.                 - linux
  126.       hostNetwork: true
  127.       priorityClassName: system-node-critical
  128.       tolerations:
  129.       - operator: Exists
  130.         effect: NoSchedule
  131.       serviceAccountName: flannel
  132.       initContainers:
  133.       - name: install-cni-plugin
  134.         image: docker.io/flannel/flannel-cni-plugin:v1.1.2
  135.        #image: docker.io/rancher/mirrored-flannelcni-flannel-cni-plugin:v1.1.2
  136.         command:
  137.         - cp
  138.         args:
  139.         - -f
  140.         - /flannel
  141.         - /opt/cni/bin/flannel
  142.         volumeMounts:
  143.         - name: cni-plugin
  144.           mountPath: /opt/cni/bin
  145.       - name: install-cni
  146.         image: docker.io/flannel/flannel:v0.21.3
  147.        #image: docker.io/rancher/mirrored-flannelcni-flannel:v0.21.3
  148.         command:
  149.         - cp
  150.         args:
  151.         - -f
  152.         - /etc/kube-flannel/cni-conf.json
  153.         - /etc/cni/net.d/10-flannel.conflist
  154.         volumeMounts:
  155.         - name: cni
  156.           mountPath: /etc/cni/net.d
  157.         - name: flannel-cfg
  158.           mountPath: /etc/kube-flannel/
  159.       containers:
  160.       - name: kube-flannel
  161.         image: docker.io/flannel/flannel:v0.21.3
  162.        #image: quay-mirror.qiniu.com/coreos/flannel:v0.21.3 
  163.        #image: docker.io/rancher/mirrored-flannelcni-flannel:v0.21.3
  164.         command:
  165.         - /opt/bin/flanneld
  166.         args:
  167.         - --ip-masq
  168.         - --kube-subnet-mgr
  169.         resources:
  170.           requests:
  171.             cpu: "100m"
  172.             memory: "50Mi"
  173.         securityContext:
  174.           privileged: false
  175.           capabilities:
  176.             add: ["NET_ADMIN", "NET_RAW"]
  177.         env:
  178.         - name: POD_NAME
  179.           valueFrom:
  180.             fieldRef:
  181.               fieldPath: metadata.name
  182.         - name: POD_NAMESPACE
  183.           valueFrom:
  184.             fieldRef:
  185.               fieldPath: metadata.namespace
  186.         - name: EVENT_QUEUE_DEPTH
  187.           value: "5000"
  188.         volumeMounts:
  189.         - name: run
  190.           mountPath: /run/flannel
  191.         - name: flannel-cfg
  192.           mountPath: /etc/kube-flannel/
  193.         - name: xtables-lock
  194.           mountPath: /run/xtables.lock
  195.       volumes:
  196.       - name: run
  197.         hostPath:
  198.           path: /run/flannel
  199.       - name: cni-plugin
  200.         hostPath:
  201.           path: /opt/cni/bin
  202.       - name: cni
  203.         hostPath:
  204.           path: /etc/cni/net.d
  205.       - name: flannel-cfg
  206.         configMap:
  207.           name: kube-flannel-cfg
  208.       - name: xtables-lock
  209.         hostPath:
  210.           path: /run/xtables.lock
  211.           type: FileOrCreate

应用flannel组件

kubectl apply -f kube-flannel.yml

查询flannel的pod信息

kubectl get pod -n kube-flannel  

查看pod和node节点

kubectl get pod -n kube-system

kubectl get node

六、部署node节点

将k8s-node01节点添加到集群中

kubeadm join 192.168.66.200:6443 --token abcdef.0123456789abcdef \
    --discovery-token-ca-cert-hash sha256:8ba059662766394bbc081324dfba0bc6a1360687f99e93a1b5c7c3a1e6d53097 

查询node节点情况

登录master01节点检查所有组件运行情况

kubectl get pod -A -owide

声明：本文内容由网友自发贡献，不代表【wpsshop博客】立场，版权归原作者所有，本站不承担相应法律责任。如您发现有侵权的内容，请联系我们。转载请注明出处：https://www.wpsshop.cn/w/煮酒与君饮/article/detail/952907
推荐阅读
相关标签

Copyright © 2003-2013 www.wpsshop.cn 版权所有，并保留所有权利。

  

闽ICP备14008679号