本地k8s环境minikube搭建过程

首先要安装docker这个环境是需要自己安装的。相关步骤如下：

1
2
3
4
5
6
7
8
9
10
11
yum install -y yum-utils device-mapper-persistent-data lvm2
yum-config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo
安装docker
yum list docker-ce --showduplicates | sort -r 查看docker相关版本
#yum install docker-ce #由于repo中默认只开启stable仓库，故这里安装的是最新稳定版。
#yum install <FQPN> # 例如：yum install docker-ce-18.06.0.ce -y
#以下是验证过的版本，建议安装
yum install docker-ce-18.06.0.ce -y
systemctl start docker
systemctl enable docker
docker version（因为安装的是1.13.4版本的k8s,建议安装docker18.06）

然后使用阿里云修改好的minikube进行安装，否则在初始化minikube的时候会卡在墙上下不来

1
2
3
curl -Lo minikube http://kubernetes.oss-cn-hangzhou.aliyuncs.com/minikube/releases/v0.35.0/minikube-linux-amd64
chmod +x minikube
mv minikube /usr/bin/minikube
注意一点要关掉swap：关闭命令swapoff -a

加载阿里云k8s的官方源并且安装相关命令组件

1
2
3
4
5
6
7
8
9
10
cd /etc/yum.repos.d/
cat>>kubernetes.repo<<EOF
[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
enabled=1
gpgcheck=1
repo_gpgcheck=1
gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg https://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
EOF
1
2
yum install kubectl kubelet kubeadm -y
systemctl start kubelet && systemctl enable kubelet

使用缺省VirtualBox驱动来创建Kubernetes本地环境

minikube start --registry-mirror=https://registry.docker-cn.com

出现如下字样

- Verifying component health .....
+ kubectl is now configured to use "minikube"
= Done! Thank you for using minikube!

则本地的minikube安装完成。当然这个不能访问外网，单独装ingress或者端口转发即可

############################################################################

----------------------------------我是分割线------------------------------------割一下--------------------------------

############################################################################

ingress安装方法：

生成ingress：

创建depolyment.yaml:

apiVersion: v1
kind: Namespace
metadata:
  name: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: nginx-configuration
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: tcp-services
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: udp-services
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: nginx-ingress-serviceaccount
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: ClusterRole
metadata:
  name: nginx-ingress-clusterrole
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
rules:
  - apiGroups:
      - ""
    resources:
      - configmaps
      - endpoints
      - nodes
      - pods
      - secrets
    verbs:
      - list
      - watch
  - apiGroups:
      - ""
    resources:
      - nodes
    verbs:
      - get
  - apiGroups:
      - ""
    resources:
      - services
    verbs:
      - get
      - list
      - watch
  - apiGroups:
      - "extensions"
    resources:
      - ingresses
    verbs:
      - get
      - list
      - watch
  - apiGroups:
      - ""
    resources:
      - events
    verbs:
      - create
      - patch
  - apiGroups:
      - "extensions"
    resources:
      - ingresses/status
    verbs:
      - update
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: Role
metadata:
  name: nginx-ingress-role
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
rules:
  - apiGroups:
      - ""
    resources:
      - configmaps
      - pods
      - secrets
      - namespaces
    verbs:
      - get
  - apiGroups:
      - ""
    resources:
      - configmaps
    resourceNames:
      # Defaults to "<election-id>-<ingress-class>"
      # Here: "<ingress-controller-leader>-<nginx>"
      # This has to be adapted if you change either parameter
      # when launching the nginx-ingress-controller.
      - "ingress-controller-leader-nginx"
    verbs:
      - get
      - update
  - apiGroups:
      - ""
    resources:
      - configmaps
    verbs:
      - create
  - apiGroups:
      - ""
    resources:
      - endpoints
    verbs:
      - get
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: RoleBinding
metadata:
  name: nginx-ingress-role-nisa-binding
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: Role
  name: nginx-ingress-role
subjects:
  - kind: ServiceAccount
    name: nginx-ingress-serviceaccount
    namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: ClusterRoleBinding
metadata:
  name: nginx-ingress-clusterrole-nisa-binding
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: nginx-ingress-clusterrole
subjects:
  - kind: ServiceAccount
    name: nginx-ingress-serviceaccount
    namespace: kube-system
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-ingress-controller
  namespace: kube-system
  labels:
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
spec:
  replicas: 1
  selector:
    matchLabels:
      app.kubernetes.io/name: ingress-nginx
      app.kubernetes.io/part-of: ingress-nginx
  template:
    metadata:
      labels:
        app.kubernetes.io/name: ingress-nginx
        app.kubernetes.io/part-of: ingress-nginx
      annotations:
        prometheus.io/port: "10254"
        prometheus.io/scrape: "true"
    spec:
      serviceAccountName: nginx-ingress-serviceaccount
      containers:
        - name: nginx-ingress-controller
          image: quay.io/kubernetes-ingress-controller/nginx-ingress-controller:0.23.0
          args:
            - /nginx-ingress-controller
            - --configmap=$(POD_NAMESPACE)/nginx-configuration
            - --tcp-services-configmap=$(POD_NAMESPACE)/tcp-services
            - --udp-services-configmap=$(POD_NAMESPACE)/udp-services
            - --publish-service=$(POD_NAMESPACE)/ingress-nginx
            - --annotations-prefix=nginx.ingress.kubernetes.io
          securityContext:
            allowPrivilegeEscalation: true
            capabilities:
              drop:
                - ALL
              add:
                - NET_BIND_SERVICE
            # www-data -> 33
            runAsUser: 33
          env:
            - name: POD_NAME
              valueFrom:
                fieldRef:
                  fieldPath: metadata.name
            - name: POD_NAMESPACE
              valueFrom:
                fieldRef:
                  fieldPath: metadata.namespace
          ports:
            - name: http
              containerPort: 80
            - name: https
              containerPort: 443
          livenessProbe:
            failureThreshold: 3
            httpGet:
              path: /healthz
              port: 10254
              scheme: HTTP
            initialDelaySeconds: 10
            periodSeconds: 10
            successThreshold: 1
            timeoutSeconds: 10
          readinessProbe:
            failureThreshold: 3
            httpGet:
              path: /healthz
              port: 10254
              scheme: HTTP
            periodSeconds: 10
            successThreshold: 1
            timeoutSeconds: 10
---

 

---

 

---

再创建svc,yaml：

Service:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
apiVersion: v1
kind: Service
metadata:
  annotations:
    #service.beta.kubernetes.io/alicloud-loadbalancer-id: "lb-wz9du18pa4e7f93vetzww"
  labels:
    app: nginx-ingress
  name: nginx-ingress
  namespace: kube-system
spec:
  ports:
  - name: http
    nodePort: 30468
    port: 80
    protocol: TCP
    targetPort: 80
  - name: https
    nodePort: 30471
    port: 443
    protocol: TCP
    targetPort: 443
  selector:
    #app: ingress-nginx
    app.kubernetes.io/name: ingress-nginx
  #type: LoadBalancer
  type: NodePort
status:
  loadBalancer:
    ingress:
    - ip: 39.108.26.119（此处更改成自己本机ip）

 

以上yaml创建pod的命令是：

kubectl apply -f xxxx.yaml

业务镜像可以拉取gitlab的，这里没做cofigmap，需要自己配。生成业务编排需自己编写yaml

以下是简单安装脚本。

#!/bin/bash
#安装docker相关，用以拉取本地所需镜像,版本采用docker-ce 18.06版，支持1.13版kubernetes
#检测网卡是否是固定ip
grep -rE "dhcp" /etc/sysconfig/network-scripts/ifcfg-*
if [ $? -eq 0 ];
then
    echo "网卡为DHCP模式请更改为规定ip"
    exit
else
    echo "网卡正常。"
fi
yum clean all && yum repolist
yum install -y yum-utils device-mapper-persistent-data lvm2
yum-config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo
yum install docker-ce-18.06.0.ce -y
systemctl start docker
systemctl enable docker
VERSION=`docker version`
if [ $? -eq 0 ];
then
    echo "输出docker版本信息：$VERSION"
else
    echo "docker安装出错，请检查错误日志"
    exit
fi
echo "1" >/proc/sys/net/bridge/bridge-nf-call-iptables #此步是保证iptables正确转发获取镜像，否则会报dns解析错误
########获取minikube二进制文件并且添加系统命令########
cd /data
curl -Lo minikube http://kubernetes.oss-cn-hangzhou.aliyuncs.com/minikube/releases/v0.35.0/minikube-linux-amd64
chmod +x minikube
mv minikube /usr/bin/minikube
swapoff -a #强制关闭swap不然初始化的时候会提示错误
cd /etc/yum.repos.d/
cat>>kubernetes.repo<<EOF
[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64/
enabled=1
gpgcheck=1
repo_gpgcheck=1
gpgkey=https://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg https://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
EOF
yum install kubectl kubelet kubeadm -y
systemctl start kubelet && systemctl enable kubelet
########启动minikube########
minikube start --vm-driver=none
if [ $? -eq 0 ];
then
    echo "minikube初始化成功"
else
    echo "minikube初始化失败,请检查报错输出，重新执行初始化命令minikube start --vm-driver=none 命令，如果仍有报错，请执行清理集群命令minikube delete，并重新执行初始化命令！"
    minikube delete
    exit
fi
#缺省Minikube使用VirtualBox驱动来创建Kubernetes本地环境
#minikube start --registry-mirror=https://registry.docker-cn.com
STATUS=`kubectl get node | awk '{print$2}' | sed -n '2p'`
if [ $STATUS = "Ready" ];
then
    echo "输出集群状态$STATUS"
else
    echo "输出状态不是Ready，请联系运维."
fi
#echo "输出集群状态$STATUS"
#echo "输出状态不是Ready，请联系运维."

 

posted @ 2019-03-29 10:21  林锅  阅读(3969)  评论(0)  编辑  收藏  举报