[kubernetes]-helm部署高可用自愈mysql集群

川长思鸟来 2022-12-11 12:21 315阅读 0赞

# helm使用operator部署高可用自愈mysql集群 #

> 部署完第一天后 vmware磁盘满了 导致有几台node崩了。。。可能截图会有ip对不上的问题。

### 安装helm ###

##### 下载安装包 #####

https://github.com/helm/helm/releases
    cd  /kubernetes/helm
    tar zxvf helm-v2.13.1-linux-amd64.tar.gz
    ln -s /kubernetes/helm/linux-amd64/helm  /usr/bin/
    helm version

##### 安装tiller #####

# 指向阿里云的仓库 Helm 默认会去 storage.googleapis.com 拉取镜像
    helm init --client-only --stable-repo-url https://aliacs-app-catalog.oss-cn-hangzhou.aliyuncs.com/charts/
    helm repo add incubator https://aliacs-app-catalog.oss-cn-hangzhou.aliyuncs.com/charts-incubator/
    helm repo update

![55efbf52b107bdc03f6ccf2b5b3f3544.png][]

# 创建TLS认证服务端 因为官方的镜像无法拉取，使用-i指定镜像
    helm init --service-account tiller --upgrade -i registry.cn-hangzhou.aliyuncs.com/google_containers/tiller:v2.13.1 --tiller-tls-cert /etc/kubernetes/ssl/tiller001.pem --tiller-tls-key /etc/kubernetes/ssl/tiller001-key.pem --tls-ca-cert /etc/kubernetes/ssl/ca.pem --tiller-namespace kube-system --stable-repo-url https://kubernetes.oss-cn-hangzhou.aliyuncs.com/charts
    # 创建serviceaccount
    kubectl create serviceaccount --namespace kube-system tiller
    # 创建角色绑定
    kubectl create clusterrolebinding tiller-cluster-rule --clusterrole=cluster-admin --serviceaccount=kube-system:tiller

##### 验证 #####

kubectl get deploy --namespace kube-system tiller-deploy -o yaml|grep serviceAccount
    kubectl -n kube-system get pods|grep tiller
    helm version

![96a44f46dfbca6ea6f52462089dd9482.png][]

### 安装Presslabs MySQL Operator ###

Presslabs MySQL Operator架构图如下

![1cd604b20cdfa4ae23cef2a97d5c2901.png][]

##### 下载相关charts到本地 #####

git clone https://github.com/kubernetes/charts.git

##### 部署Presslabs MySQL Operator #####

cd  mysql-operator/charts/mysql-operator/
     vi values.yaml

![b445307b5cc6a150bc032d8bb2046d35.png][]

##### 创建rabc-1.yaml #####

---
    # 唯一需要修改的地方只有namespace,根据实际情况定义
    apiVersion: v1
    kind: ServiceAccount
    metadata:
      name: nfs-client-provisioner
      # replace with namespace where provisioner is deployed
      namespace: operator        #根据实际环境设定namespace,下面类同
    ---
    kind: ClusterRole
    apiVersion: rbac.authorization.k8s.io/v1
    metadata:
      name: nfs-client-provisioner-runner
      namespace: operator
    rules:
      - apiGroups: [""]
        resources: ["persistentvolumes"]
        verbs: ["get", "list", "watch", "create", "delete"]
      - apiGroups: [""]
        resources: ["persistentvolumeclaims"]
        verbs: ["get", "list", "watch", "update"]
      - apiGroups: ["storage.k8s.io"]
        resources: ["storageclasses"]
        verbs: ["get", "list", "watch"]
      - apiGroups: [""]
        resources: ["events"]
        verbs: ["create", "update", "patch"]
    ---
    kind: ClusterRoleBinding
    apiVersion: rbac.authorization.k8s.io/v1
    metadata:
      name: run-nfs-client-provisioner
    subjects:
      - kind: ServiceAccount
        name: nfs-client-provisioner
        # replace with namespace where provisioner is deployed
        namespace: operator
    roleRef:
      kind: ClusterRole
      name: nfs-client-provisioner-runner
      apiGroup: rbac.authorization.k8s.io
    ---
    kind: Role
    apiVersion: rbac.authorization.k8s.io/v1
    metadata:
      name: leader-locking-nfs-client-provisioner
        # replace with namespace where provisioner is deployed
      namespace: operator
    rules:
      - apiGroups: [""]
        resources: ["endpoints"]
        verbs: ["get", "list", "watch", "create", "update", "patch"]
    ---
    kind: RoleBinding
    apiVersion: rbac.authorization.k8s.io/v1
    metadata:
      name: leader-locking-nfs-client-provisioner
    subjects:
      - kind: ServiceAccount
        name: nfs-client-provisioner
        # replace with namespace where provisioner is deployed
        namespace: operator
    roleRef:
      kind: Role
      name: leader-locking-nfs-client-provisioner
      apiGroup: rbac.authorization.k8s.io

##### 创建nfs-StorageClass-2.yaml #####

apiVersion: storage.k8s.io/v1
    kind: StorageClass
    metadata:
      name: presslabs-managed-nfs-storage
      namespace: operator
    provisioner: presslabs-mysql-nfs-storage #这里的名称要和provisioner配置文件中的环境变量PROVISIONER_NAME保持一致parameters:  archiveOnDelete: "false"

##### helm安装presslabs/mysql-operator #####

# 添加源
    helm repo add presslabs https://presslabs.github.io/charts
    
    # 报错Error: failed to download "presslabs-mysql-operator" (hint: running `helm repo update` may help)
    # helm repo update
    
    # helm安装
    # helm install presslabs/mysql-operator   -f values.yaml --namespace operator
    helm install presslabs/mysql-operator --name presslabs  -f values.yaml --namespace operator --set rbac.enabled=true

![a12c3f04956183a35eef0c7faafcb12c.png][]

##### 调整values.yaml里replicas的值为3 #####

# helm删除
    # helm delete presslabs  --purge
    # 更新value
    helm upgrade presslabs --values values.yaml presslabs/mysql-operator --namespace operator --set rbac.enabled=true

![6d4404b57cacd81ba681db02c4724267.png][]

![afd45604306777926cccf237c5331681.png][]

##### 查看leader #####

# 需要看每个pod的日志  我这边是持久化了 看到被选举为leader  而刚好这个第二个pod 获得了2票
    # 直接查看日志命令
    kubectl log -f presslabs-mysql-operator-2 -c orchestrator -n operator

![910f4f234078fc23543e92c3382c5721.png][]

![f66665dd79261b0c892eb2603ea25837.png][]

> 里面的operator容器和orchestrator容器都会有选主的过程，这里对应的应该是presslabs-mysql-operator-2， Operator POD在起来之后，其中给一个会成为leader leader的ip是svc的ip 会出现如下字样
> 
> 2020/09/30 01:37:59 \[INFO\] raft: Election won. Tally: 2  
> 2020/09/30 01:37:59 \[INFO\] raft: Node at 10.109.138.165:10008 \[Leader\] entering Leader state

# 查看对应svc的ip
    kubectl get svc  -o wide  -n operator

![7c77c0c32d7ba0aadf33197564558a77.png][]

##### 创建ingress ingress-4.yaml #####

apiVersion: extensions/v1beta1
    kind: Ingress
    metadata:
    # 通过添加下面的annotations 来开启白名单
    # 关闭80强制跳转443 为ingress配置增加注解（annotations）：nginx.ingress.kubernetes.io/ssl-redirect: 'false' 就可以禁止http强制跳转至https
      annotations:
        #nginx.ingress.kubernetes.io/whitelist-source-range: "60.191.70.64/29, xx.xxx.0.0/16"
        nginx.ingress.kubernetes.io/ssl-redirect: 'false'
      name: presslab-mysql-operator-ingress
      namespace: operator
    spec:
      tls:
       - hosts:
         - test-mysql.jiaminxu.com
    #     secretName: ihaozhuo
      rules:
        - host: test-mysql.jiaminxu.com
          http:
            paths:
            - path: /
              backend:
                serviceName: presslabs-mysql-operator
                servicePort: 80

![4daad03630b032cce9a76cd00e9e0f7f.png][]

##### 访问自定义的域名 #####

![eb66aaea78a022779261ce5d6ee16fe7.png][]

##### 部署Presslabs MySQL Cluster #####

# cd -
    cd  mysql-operator/charts/mysql-cluster/
    # 修改repilicas的值为3 部署一个3节点的percona mysql集群  并且账号密码之类的需要自行设置！
    vi values.yaml

helm install presslabs/mysql-cluster --name my-cluster  -f values.yaml --namespace operator --set rbac.enabled=true
    # helm delete my-cluster  --purge

![c76075c106fd7f4f9ed6399e0bd78977.png][]

![97d87305cae1b56abaf610d74d9e5db1.png][]

![9fc8f4682669f300b352824d38153fe1.png][]

##### 查看svc #####

# 查看svc
    kubectl get svc -n operator

![ed444579c02a4e203d1f1491eb31ecc4.png][]

在节点部署之后，会借助orchestractor进行leader选举，其中my-cluster-mysql-cluster-db-mysql-master相当于当前选举出来的的master节点，可以进行读写操作：

kubectl describe svc my-cluster-mysql-cluster-db-mysql-master  -n operator

![de2f79d5e73deef16f5b0bb65ea39397.png][]

而my-cluster-mysql-cluster-db-mysql里面是所有的mysql节点，可以用于读操作： 看了ip包含master的ip

kubectl describe svc my-cluster-mysql-cluster-db-mysql  -n operator

![f652144bb54de0aad9c0274e6b08aa61.png][]

##### 测试高可用性 #####

如果operator/orchestrator发生故障，如果又恰好是leader，则会重新进行leader选举，比如我将现在的leader presslabs-mysql-operator-2 POD重启，则presslabs-mysql-operator-1成为了新的leader：

![1b20414aafc1ad53735703d5c77a4abb.png][]

![7c77c0c32d7ba0aadf33197564558a77.png][]

> 理论上整个过程mysql业务集群没有任何感知和影响。 这个感知我没有测试

删除mysql master的pod测试 如图 ip发生了改变

![1f85c26addaee28ce085943af2b57fbe.png][]

也可以在网页控制台调整 类似如下

![2bf13d778a4c19e9f92de7edb1a03f6d.png][]

![0ea908282ca19d301e946d10904c1ea9.png][]

此时这个原先的master节点会变成stop 状态 需要点开 ⚙️ start slave

![447f09105861b39fb6866088060db81f.png][]

##### 测试删除pod #####

kubectl delete pod my-cluster-mysql-cluster-db-mysql-2 -n operator

![00cb64ffe6fb774782933367944b35a7.png][]

![8d76f61d0dae5c0197de7b66e7735822.png][]

![24bfa06bb7f4dea6e6de4dd67bbbc4c1.png][]

提示master not replicating

这里点开master的⚙️ 然后reset slave 就好了

![0719ba1deb34111f0be2ab45f5c25992.png][]

![a66167d7a1cd89756038ef0dfda5ff44.png][]

还在摸索阶段 如果有错误 欢迎指正

参考

https://blog.csdn.net/cloudvtech/article/details/105649723

[55efbf52b107bdc03f6ccf2b5b3f3544.png]: /images/20221123/f41b6742840941b1bc4724758200392b.png
[96a44f46dfbca6ea6f52462089dd9482.png]: /images/20221123/2ce721885e27403eba6bb80340318baf.png
[1cd604b20cdfa4ae23cef2a97d5c2901.png]: /images/20221123/6f06a51f00ab4392879ee60a51a99370.png
[b445307b5cc6a150bc032d8bb2046d35.png]: /images/20221123/73863f48b6464c71a62fe1d1d78db69e.png
[a12c3f04956183a35eef0c7faafcb12c.png]: /images/20221123/9657886cad61443887865007b1202cc9.png
[6d4404b57cacd81ba681db02c4724267.png]: /images/20221123/c2490f33cb3c47abaf7d13ec6225f373.png
[afd45604306777926cccf237c5331681.png]: /images/20221123/ee4ffb308d7347818af1d7f0decbd4f6.png
[910f4f234078fc23543e92c3382c5721.png]: /images/20221123/dc4d6f82f8bc43b1a2dbff91a12d209a.png
[f66665dd79261b0c892eb2603ea25837.png]: /images/20221123/7f58e16781194a54b293ad355f0b034b.png
[7c77c0c32d7ba0aadf33197564558a77.png]: /images/20221123/55a0a78f9841409192593dacaba8390a.png
[4daad03630b032cce9a76cd00e9e0f7f.png]: /images/20221123/1fb73202557e4631a574d15752d99849.png
[eb66aaea78a022779261ce5d6ee16fe7.png]: /images/20221123/8f033159982b4ce2a59ce64b0e9f9d69.png
[c76075c106fd7f4f9ed6399e0bd78977.png]: /images/20221123/baabfdff1fb1476b813adce507794471.png
[97d87305cae1b56abaf610d74d9e5db1.png]: /images/20221123/1252f540069241d29a1bcc1e73cd79c4.png
[9fc8f4682669f300b352824d38153fe1.png]: /images/20221123/1ce843d031564b7b9d4a71107d599cfc.png
[ed444579c02a4e203d1f1491eb31ecc4.png]: /images/20221123/59c7a2607d634c81b9a501c552fff62a.png
[de2f79d5e73deef16f5b0bb65ea39397.png]: /images/20221123/0d33e47bc8034b909f4b878c192487ee.png
[f652144bb54de0aad9c0274e6b08aa61.png]: /images/20221123/69f49106865e4ec2b37813c9e3c1995d.png
[1b20414aafc1ad53735703d5c77a4abb.png]: /images/20221123/badf61b8a7bd4daaaaa238b2100af771.png
[1f85c26addaee28ce085943af2b57fbe.png]: /images/20221123/8fba55fa04fe4d0a948af32285dab049.png
[2bf13d778a4c19e9f92de7edb1a03f6d.png]: /images/20221123/b10f9a055e5946cfabed056fdedff3b1.png
[0ea908282ca19d301e946d10904c1ea9.png]: /images/20221123/4fb0ce9f5caf42fd85d613516a52fa86.png
[447f09105861b39fb6866088060db81f.png]: /images/20221123/a0b14da1afd64e27bd206b1a25d18e05.png
[00cb64ffe6fb774782933367944b35a7.png]: /images/20221123/1bcc347af9f04012ac8ddf2ad3e79163.png
[8d76f61d0dae5c0197de7b66e7735822.png]: /images/20221123/5d829749ccf0461cae52932b5b5fe37d.png
[24bfa06bb7f4dea6e6de4dd67bbbc4c1.png]: /images/20221123/9ad4e90170b0471d86478c4d495ef984.png
[0719ba1deb34111f0be2ab45f5c25992.png]: /images/20221123/d4d9ee41f3a1492dbbc10ff60fc6abc7.png
[a66167d7a1cd89756038ef0dfda5ff44.png]: /images/20221123/da27d2cb57204e0d92f6f47cc9cc10b8.png