基于StatefulSet控制器运行MySQL一主多从

大洋百度

已于 2023-02-09 19:17:05 修改

阅读量433

点赞数 1

文章标签：缓存 redis 运维

于 2023-01-05 09:12:08 首次发布

本文链接：https://blog.csdn.net/weixin_46476452/article/details/128539569

版权

1、制作configmap资源，用于提供mysql的配置

2、创建service资源，用于定义数据库如何访问

3、制作statefulset资源，用于控制pod

四、验证数据一致性

1、数据同步实践

五、查验数据是否存在

简介：

众所周知pod控制器是K8S集群重要的资源，用于控制pod；Pod通过控制器实现容器的运维，如伸缩、升级等，控制器决定了创建pod资源的方式和类型

pod控制器分为两类：1、有状态应用控制器 deployment 2、无状态应用控制器statefulse

无状态应用控制器常用的deployment，在deployment控制下的pod都是一样的，没有序列关系；

statefulse与deployment不同的是，它会让每个容器有自己的唯一标识符，多用于数据库服务，因为数据库服务一般有主从关系，必须按照顺序进行扩缩容、删除和终止，进行有序的更新

有状态应用特点：

1）实例之间有差别，每个实例都有自己的独特性，元数据不同，例如etcd，zookeeper
2）实例之间不对等的关系，以及依靠外部存储的应用。

无状态应用特点：

1）deployment 认为所有的pod都是一样的
2）不用考虑顺序的要求
3）不用考虑在哪个node节点上运行
4）可以随意扩容和缩容

一、基础镜像准备

1、准备xtrabackup镜像

下载镜像
# nerdctl pull registry.cn-hangzhou.aliyuncs.com/hxpdocker/xtrabackup:1.0

修改tag
# nerdctl tag registry.cn-hangzhou.aliyuncs.com/hxpdocker/xtrabackup:1.0 harbor.magedu.net/baseimages/xtrabackup:1.0

上传到本地harbor
# nerdctl push  harbor.magedu.net/baseimages/xtrabackup:1.0

2、准备mysql镜像

# nerdctl pull mysql:5.7

进入mysql容器中查看下mysql具体版本号

修改镜像Tag, 并上传至harbor

root@master1:~# nerdctl tag mysql:5.7 harbor.magedu.net/baseimages/mysql:5.7.40
root@master1:~# nerdctl push harbor.magedu.net/baseimages/mysql:5.7.40

二、创建PV

创建pv资源

root@master1:/mysql/pv# vim mysql-persistentvolume.yaml
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-1
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-1
    server: 172.31.7.109
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-2
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-2
    server: 172.31.7.109
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-3
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-3
    server: 172.31.7.109
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-4
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-4
    server: 172.31.7.109
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-5
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-5
    server: 172.31.7.109

---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: mysql-datadir-6
  namespace: magedu
spec:
  capacity:
    storage: 50Gi
  accessModes:
    - ReadWriteOnce
  nfs:
    path: /data/k8sdata/magedu/mysql-datadir-6
    server: 172.31.7.109

上面yaml文件中的6个nfs共享目录需要在存储服务器上创建

1、安装nfs服务端
root@haproxy02:~# apt install nfs-server

2、创建共享目录
root@haproxy02:~# mkdir -p /data/k8sdata/magedu/mysql-datadir-{1,2,3,4,5,6}

3、给目录赋权
root@haproxy02:~# cat /etc/exports  | grep /data/k8sdata
/data/k8sdata *(rw,no_root_squash)

root@haproxy02:~# exportfs -arv

4、重启服务并设置开机自启
root@haproxy02:~# systemctl restart nfs-server && systemctl enable nfs-server

创建pv资源

root@master1:/mysql/pv# kubectl apply -f mysql-persistentvolume.yaml
persistentvolume/mysql-datadir-1 created
persistentvolume/mysql-datadir-2 created
persistentvolume/mysql-datadir-3 created
persistentvolume/mysql-datadir-4 created
persistentvolume/mysql-datadir-5 created
persistentvolume/mysql-datadir-6 created

三、创建各种yaml资源

1、制作configmap资源，用于提供mysql的配置

root@master1:/mysql# vim mysql-configmap.yaml
apiVersion: v1
kind: ConfigMap
metadata:
  name: mysql
  namespace: magedu
  labels:
    app: mysql
data:          #data下有两段配置，一个是master，一个是slave
  master.cnf: |
    # Apply this config only on the master.
    [mysqld]
    log-bin       #master开启了binlog
    log_bin_trust_function_creators=1
    lower_case_table_names=1
  slave.cnf: |
    # Apply this config only on slaves.
    [mysqld]
    super-read-only             #slave是只读的，不能写入数据
    log_bin_trust_function_creators=1

root@master1:/mysql# kubectl apply -f mysql-configmap.yaml
configmap/mysql created

root@master1:/mysql# kubectl get configmap -n magedu
NAME               DATA   AGE
mysql              2      16s

2、创建service资源，用于定义数据库如何访问

root@master1:/mysql# vim mysql-services.yaml
apiVersion: v1
kind: Service
metadata:
  namespace: magedu
  name: mysql
  labels:
    app: mysql
spec:
  ports:
  - name: mysql
    port: 3306
  clusterIP: None  #无头服务，可通过pod名称访问数据库服务
  selector:
    app: mysql

---
# Client service for connecting to any MySQL instance for reads.
# For writes, you must instead connect to the master: mysql-0.mysql.

apiVersion: v1
kind: Service
metadata:
  name: mysql-read #通过service访问服务
  namespace: magedu
  labels:
    app: mysql
spec:
  ports:
  - name: mysql
    port: 3306
  selector:
    app: mysql

root@master1:/mysql# kubectl apply -f mysql-configmap.yaml

root@master1:/mysql# kubectl get svc -n magedu
NAME               TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)          AGE
mysql              ClusterIP   None             <none>        3306/TCP         62s
mysql-read         ClusterIP   10.100.254.189   <none>        3306/TCP         62s

3、制作statefulset资源，用于控制pod

root@master1:/mysql# vim mysql-statefulset.yaml
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: mysql
  namespace: magedu
spec:
  selector:
    matchLabels:
      app: mysql
  serviceName: mysql
  replicas: 2
  template:
    metadata:
      labels:
        app: mysql
    spec:
      initContainers:
      - name: init-mysql #初始化容器1、基于当前pod name匹配角色是master还是slave，master只有一个就是pod0,其它都是slave,并动态生成相对应的配置文件，如果是master就加载master的配置文件，如果是slave就加载slave的配置文件
        image: harbor.magedu.net/baseimages/mysql:5.7.40
        command:
        - bash
        - "-c"
        - |
          set -ex
          # Generate mysql server-id from pod ordinal index.
          [[ `hostname` =~ -([0-9]+)$ ]] || exit 1 #匹配hostname的最后一位、最后是一个顺序叠加的整数，看最后一个值是不是0
          ordinal=${BASH_REMATCH[1]}
          echo [mysqld] > /mnt/conf.d/server-id.cnf
          # Add an offset to avoid reserved server-id=0 value.
          echo server-id=$((100 + $ordinal)) >> /mnt/conf.d/server-id.cnf
          # Copy appropriate conf.d files from config-map to emptyDir.
          if [[ $ordinal -eq 0 ]]; then #如果是master、则cpmaster配置文件
            cp /mnt/config-map/master.cnf /mnt/conf.d/
          else #否则cp slave配置文件
            cp /mnt/config-map/slave.cnf /mnt/conf.d/
          fi
        volumeMounts:
        - name: conf #临时卷、emptyDir
          mountPath: /mnt/conf.d
        - name: config-map
          mountPath: /mnt/config-map
      - name: clone-mysql #初始化容器2、用于生成mysql配置文件、并从上一个pod完成首次的全量数据clone(slave 3从slave2 clone，而不是每个slave都从master clone实现首次全量同步,但是后期都是与master实现增量同步)
        image: harbor.magedu.net/baseimages/xtrabackup:1.0
        command:
        - bash
        - "-c"
        - |
          set -ex
          # Skip the clone if data already exists.
          [[ -d /var/lib/mysql/mysql ]] && exit 0
          # Skip the clone on master (ordinal index 0).
          [[ `hostname` =~ -([0-9]+)$ ]] || exit 1
          ordinal=${BASH_REMATCH[1]}
          [[ $ordinal -eq 0 ]] && exit 0 #如果最后一位是0(master)则退出clone过程
          # Clone data from previous peer.
          ncat --recv-only mysql-$(($ordinal-1)).mysql 3307 | xbstream -x -C /var/lib/mysql #从上一个pod执行clone(binlog),xbstream为解压缩命令
          # Prepare the backup.xue
          xtrabackup --prepare --target-dir=/var/lib/mysql #通过xtrabackup恢复binlog
        volumeMounts:
        - name: data
          mountPath: /var/lib/mysql
          subPath: mysql
        - name: conf
          mountPath: /etc/mysql/conf.d
      containers:
      - name: mysql #业务容器1(mysql主容器)
        image: harbor.magedu.net/baseimages/mysql:5.7.40
        env:
        - name: MYSQL_ALLOW_EMPTY_PASSWORD
          value: "1"
        ports:
        - name: mysql
          containerPort: 3306
        volumeMounts:
        - name: data #挂载数据目录至/var/lib/mysql
          mountPath: /var/lib/mysql
          subPath: mysql
        - name: conf #配置文件/etc/mysql/conf.d
          mountPath: /etc/mysql/conf.d
        resources: #资源限制
          requests:
            cpu: 500m
            memory: 1Gi
        livenessProbe: #存活探针
          exec:
            command: ["mysqladmin", "ping"]
          initialDelaySeconds: 30
          periodSeconds: 10
          timeoutSeconds: 5
        readinessProbe: #就绪探针
          exec:
            # Check we can execute queries over TCP (skip-networking is off).
            command: ["mysql", "-h", "127.0.0.1", "-e", "SELECT 1"]
          initialDelaySeconds: 5
          periodSeconds: 2
          timeoutSeconds: 1
      - name: xtrabackup #业务容器2(xtrabackup),用于后期同步master 的binglog并恢复数据
        image: harbor.magedu.net/baseimages/xtrabackup:1.0
        ports:
        - name: xtrabackup
          containerPort: 3307
        command:
        - bash
        - "-c"
        - |
          set -ex
          cd /var/lib/mysql
          # Determine binlog position of cloned data, if any.
          if [[ -f xtrabackup_slave_info ]]; then
            # XtraBackup already generated a partial "CHANGE MASTER TO" query
            # because we're cloning from an existing slave.
            mv xtrabackup_slave_info change_master_to.sql.in
            # Ignore xtrabackup_binlog_info in this case (it's useless).
            rm -f xtrabackup_binlog_info
          elif [[ -f xtrabackup_binlog_info ]]; then
            # We're cloning directly from master. Parse binlog position.
            [[ `cat xtrabackup_binlog_info` =~ ^(.*?)[[:space:]]+(.*?)$ ]] || exit 1
            rm xtrabackup_binlog_info
            echo "CHANGE MASTER TO MASTER_LOG_FILE='${BASH_REMATCH[1]}',\
                  MASTER_LOG_POS=${BASH_REMATCH[2]}" > change_master_to.sql.in #生成CHANGE MASTER命令
          fi
          # Check if we need to complete a clone by starting replication.
          if [[ -f change_master_to.sql.in ]]; then
            echo "Waiting for mysqld to be ready (accepting connections)"
            until mysql -h 127.0.0.1 -e "SELECT 1"; do sleep 1; done
            echo "Initializing replication from clone position"
            # In case of container restart, attempt this at-most-once.
            mv change_master_to.sql.in change_master_to.sql.orig
            #执行CHANGE MASTER操作并启动SLAVE
            mysql -h 127.0.0.1 <<EOF
          $(<change_master_to.sql.orig),
            MASTER_HOST='mysql-0.mysql',
            MASTER_USER='root',
            MASTER_PASSWORD='',
            MASTER_CONNECT_RETRY=10;
          START SLAVE;
          EOF
          fi
          # Start a server to send backups when requested by peers. #监听在3307端口，用于为下一个pod同步全量数据
          exec ncat --listen --keep-open --send-only --max-conns=1 3307 -c \
            "xtrabackup --backup --slave-info --stream=xbstream --host=127.0.0.1 --user=root"
        volumeMounts:
        - name: data
          mountPath: /var/lib/mysql
          subPath: mysql
        - name: conf
          mountPath: /etc/mysql/conf.d
        resources:
          requests:
            cpu: 100m
            memory: 100Mi
      volumes:
      - name: conf
        emptyDir: {}
      - name: config-map
        configMap:
          name: mysql
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: ["ReadWriteOnce"]
      resources:
        requests:
          storage: 10Gi

root@master1:/mysql# kubectl apply -f mysql-statefulset.yaml
显示报错，无非成功创建，排查五问题，可能是mysql镜像的问题，把mysql镜像换成5.7.36就可以了

重新下载个5.7.36版本的镜像，再上传到harbor

把yaml文件中mysql镜像版本改成5.7.36

image: harbor.magedu.net/baseimages/mysql:5.7.36

镜像修改后重新创建资源

查看下logs日志，显示mysqld: ready for connections就是正常状态的

root@master1:~# kubectl logs -f mysql-1 -n magedu

四、验证数据一致性

副本扩容

把2个副本调整到3个副本

创建资源，有三个pod,mysql-0是主库，其余是备库

1、数据同步实践

1.1进入主库新建一个库

1.2 进入从库可同步刚刚主库创建的数据

1.3 可以查看下同步状态，显示IO线程和SQL线程都是yes状态

1.4 手动删除mysql-0主库，验证数据是否会丢失，是否会进行同步到从库

root@master1:/mysql# kubectl delete pod mysql-0 -n magedu

1.5 删除后控制器又重新启动了一个新的mysql-0 pod,可以看见创建的时间是不一样的

五、查验数据是否存在

现在的主库mysql-0是新建的，那么之前新建的数据还在吗？

1、进入主库容器可以看见数据还是在的

root@master1:/mysql# kubectl exec -it mysql-0 bash -n magedu

2、进入从库查看数据也都是同步的，状态也都是正常的

root@master1:/mysql# kubectl exec -it mysql-1 bash -n magedu

3、进入mysql-0主库再创建一个库，看看从库是否会同步数据

root@master1:~# kubectl exec -it mysql-0 bash -n maged
root@mysql-0:/# mysql
mysql> create database n71;
Query OK, 1 row affected (0.01 sec)

root@master1:~# kubectl exec -it mysql-1 bash -n magedu
root@mysql-1:/# mysql
mysql> show databases;
+------------------------+
| Database               |
+------------------------+
| information_schema     |
| mysql                  |
| n70                    |
| n71                    |
| performance_schema     |
| sys                    |
| xtrabackup_backupfiles |

数据同步的原因在于存储卷

可以在容器中看见挂载的存储，和PV的对应关系，mysql-datadir-1对应的是绑定在data-mysql-1这个PVC上的；只要PVC的数据没删除，那么无论容器重建在哪个主机上数据都会被挂载过来