部署 MySQL
配置 mysql configmap
--- apiVersion: v1 data: mysqld.cnf: |- [mysqld] pid-file = /var/run/mysqld/mysqld.pid socket = /var/run/mysqld/mysqld.sock datadir = /var/lib/mysql bind-address = 0.0.0.0 server_id = 1918 default-time-zone='+08:00' # Disabling symbolic-links is recommended to prevent assorted security risks symbolic-links=0 max_connections=1024 default_storage_engine=innodb skip_external_locking lower_case_table_names=1 skip_host_cache skip_name_resolve character_set_server=utf8mb4 max_allowed_packet = 12M sql_mode='STRICT_TRANS_TABLES,NO_ZERO_IN_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION' # ---------- binlog & errorlog & slowlog ---------- # log_bin = mysql-bin binlog_format = ROW max_binlog_size = 100m binlog_cache_size = 4m max_binlog_cache_size = 512m expire_logs_days = 7 log-error = /var/lib/mysql/error.log slow-query-log = on slow_query_log_file = /var/lib/mysql/slow.log long_query_time = 1 [client] default-character-set=utf8mb4 kind: ConfigMap metadata: annotations: labels: app: mysql name: mysql-cm namespace: bigdata
配置 mysql svc
--- apiVersion: v1 kind: Service metadata: annotations: labels: app: mysql name: mysql-svc namespace: bigdata spec: ports: - name: mysql port: 3306 protocol: TCP targetPort: 3306 selector: app: mysql
配置 mysql deployment
mysql 的数据目录使用的 hostpath 的方式，通过亲和性来绑定节点，需要提前给对应的节点打上
mysql=这样的标签
kubectl label node <nodename> mysql=
--- apiVersion: apps/v1 kind: Deployment metadata: annotations: labels: app: mysql name: mysql namespace: bigdata spec: replicas: 1 selector: matchLabels: app: mysql template: metadata: labels: app: mysql spec: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: mysql operator: Exists containers: - name: mysql env: - name: APP_NAME value: mysql - name: NODE_NAME valueFrom: fieldRef: fieldPath: spec.nodeName - name: POD_NAME valueFrom: fieldRef: fieldPath: metadata.name - name: NAMESPACE valueFrom: fieldRef: fieldPath: metadata.namespace - name: MYSQL_ROOT_USER value: root - name: MYSQL_ROOT_PASSWORD value: YWJjMTIzCg== - name: TZ value: Asia/Shanghai - name: LANG value: en_US.UTF-8 image: mysql:5.7.31 imagePullPolicy: IfNotPresent livenessProbe: exec: command: - /bin/bash - -c - mysql -u${MYSQL_ROOT_USER} -p${MYSQL_ROOT_PASSWORD} -e 'SELECT 1' initialDelaySeconds: 30 periodSeconds: 30 timeoutSeconds: 10 ports: - containerPort: 3306 name: mysql protocol: TCP readinessProbe: exec: command: - /bin/bash - -c - mysql -u${MYSQL_ROOT_USER} -p${MYSQL_ROOT_PASSWORD} -e 'SELECT 1' initialDelaySeconds: 30 periodSeconds: 30 timeoutSeconds: 10 volumeMounts: - mountPath: /var/lib/mysql name: mysql-data-dir - mountPath: /etc/mysql/conf.d/ name: mysql-config dnsPolicy: ClusterFirst restartPolicy: Always securityContext: {} terminationGracePeriodSeconds: 10 volumes: - name: mysql-data-dir hostPath: path: /data/k8s_data/mysql type: DirectoryOrCreate - name: mysql-config configMap: name: mysql-cm
配置 mysql cronjob
- 和 deployment 的一样，需要给节点打上
mysql-back=的标签- 注意
schedule的时间是和controller-manager组件的时区一样的，如果是 kubeadm 部署的，就要注意容器内的时间是不是和本地相差八小时--- apiVersion: batch/v1 kind: CronJob metadata: name: mysql-backup namespace: bigdata spec: schedule: "0 1 * * *" successfulJobsHistoryLimit: 3 failedJobsHistoryLimit: 3 jobTemplate: spec: template: spec: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: mysql-back operator: Exists containers: - name: mysql-backup imagePullPolicy: IfNotPresent image: mysql:5.7.31 env: - name: MYSQL_ROOT_USER value: root - name: MYSQL_ROOT_PASSWORD value: YWJjMTIzCg== - name: MYSQL_HOST value: mysql-svc.bigdata.svc.cluster.local - name: TZ value: Asia/Shanghai - name: LANG value: en_US.UTF-8 command: - /bin/sh - -c - | set -ex mysqldump --host=${MYSQL_HOST} --user=${MYSQL_ROOT_USER} \ --password=${MYSQL_ROOT_PASSWORD} \ --routines --all-databases --single-transaction \ > /mysql-backup/mysql-$(date +"%Y%m%dT%H_%M_%S").sql && find /mysql-backup/ -type f -mtime +3 -exec rm -rf {} \; volumeMounts: - name: mysql-backup mountPath: /mysql-backup restartPolicy: OnFailure volumes: - name: mysql-backup hostPath: path: /data/k8s_data/mysql-back type: DirectoryOrCreate
配置 mysql-web
- phpmyadmin-github
- phpmyadmin-docker-github
- 这个是 mysql 的图形化界面
配置 mysql-web svc
- 如果有 ingress，可以配置成 clusterip 的模式
- 如果没有 ingress，可以配置成 nodeport 的模式
- 下面两个都配置了，可以各取所需
- 下面的 nodeport 配置了 30003 端口，需要检查一下自己本地有没有被占用，也可以自己随意修改，端口范围值一般是 30000-32767，具体要看 apiserver 的配置
--- apiVersion: v1 kind: Service metadata: namespace: bigdata name: phpmyadmin labels: app: phpmyadmin spec: ports: - port: 80 targetPort: 80 protocol: TCP name: http selector: app: phpmyadmin --- apiVersion: v1 kind: Service metadata: namespace: bigdata name: phpmyadmin-np labels: app: phpmyadmin spec: type: NodePort ports: - port: 80 targetPort: 80 protocol: TCP name: http-np nodePort: 30003 selector: app: phpmyadmin
配置 mysql-web deployment
- 下面的 PMA_HOST 需要替换成自己的 mysql 地址
- 下面的 PMA_PORT 需要替换成自己的 mysql 端口
--- apiVersion: apps/v1 kind: Deployment metadata: annotations: labels: app: phpmyadmin name: phpmyadmin namespace: bigdata spec: replicas: 1 selector: matchLabels: app: phpmyadmin template: metadata: annotations: labels: app: phpmyadmin spec: containers: - name: phpmyadmin image: phpmyadmin/phpmyadmin:latest imagePullPolicy: IfNotPresent ports: - containerPort: 80 env: - name: PMA_HOST value: mysql-svc.bigdata.svc.cluster.local - name: PMA_PORT value: "3306"
pod 变成 running 之后，就可以通过 ip 加 nodeport 的方式来访问了，使用 mysql 数据库内的用户名和密码登录，可以自己先提前在 mysql 数据库里面创建好
